Abstract Background: Understanding the tumor immune microenvironment (TIME) is essential for advancing cancer research and improving treatment strategies. Multiplex immunofluorescence (mIF) is a spatial proteomics imaging technique enabling simultaneous analysis of multiple markers in preserved tissues. However, mIF-derived cell abundance data pose statistical challenges, such as zero-inflation, over-dispersion, hierarchical cell relationships, and repeated measures, that must be addressed to extract meaningful insights and enhance translational impact. Methods: We developed a novel Bayesian multi-cell type analysis model that simultaneously models the relationship of immune cell abundances with clinical and epidemiological factors, while incorporating the biological relationships between immune cell populations. We applied this model to three large studies assessing the TIME of high-grade serous ovarian cancer: Nurses’ Health Study I/II (NHSI/II) (N=321), African American Cancer Epidemiology Study (AACES) (N=92), and University of Colorado Ovarian Cancer Study (UCOCS) (N=103). The mIF staining for these studies was performed using the AKOYA Biosciences OPALTM 7-Color Automation IHC Kit with the Vectra®3 Automated Quantitative Pathology Imaging System (0.499µm/pixel) utilized for image collection. InForm and HALO were utilized for spectral unmixing and cell phenotyping, respectively. Our analysis examined associations between immune cell infiltration (T-cells, B-cells, macrophages) and clinical variables (cancer stage, age at diagnosis, debulking status) with comparisons to the single-cell type model. Results: In the NHSI/II analysis, our multi-cell type model detected a positive association between age at diagnosis and abundance levels of 6 of the 7 cell types in the analysis while the single-cell type model only detected 2 of the 7. This indicates improved association detection, with our multi-cell type model. We also observed that our multi-cell type model had narrower credible intervals (CIs) for all 7 cell types demonstrating higher accuracy in the association estimation. With cancer stage as the predictor in the NHSI/II analysis, neither model detected an association although our multi-cell type model had narrower CIs for 3 of the 7 cell types compared to the single-cell type model. Despite not capturing any associations between the predictors (age, stage, debulking status) and immune cell populations in the AACES or UCOCS, our Bayesian multi-cell type model had narrower CIs for every cell type in both studies for each predictor. Discussion: Our Bayesian multi-cell type model offers a flexible framework for incorporating immune cell relationships and is well-suited for cancer studies of the TIME utilizing TMAs, regions of interest, or whole-slide imaging data. Citation Format: Chase Sakitis, Jose Laborde, Julia Wrobel, Alex C. Soupir, Christelle M. Colin-Leitzinger, Benjamin G. Bitler, Mary K. Townsend, Andrew B. Lawson, Joellen M. Schildkraut, Shelley S. Tworoger, Kathryn L. Terry, Lauren C. Peres, Brooke L. Fridley. Multi-cell type model for analyzing spatial single-cell protein imaging data with application to ovarian cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 6851.

Abstract Background: Researchers can study both the abundance and spatial architecture of cell types within the tumor microenvironment (TME) using spatial technologies. Often, Ripley’s K or nearest-neighbor G are used to measure spatial clustering of cells. These measures can be computed at various radii to assess clustering at different spatial ranges. We propose the use of functional principal component analysis (FPCAs) to model the association of the spatial clustering of T cell populations in the TME with survival from high grade serous ovarian cancer (HGSOC). Methods: We applied FPCA to study the clustering of CD3+ and CD3+CD8+ cells in the ovarian TME with survival. Five ovarian cancer studies were included in the analysis: Nurses’ Health Study (N=239), Nurses’ Health Study II (N=68), New England Case Control Study of Ovarian Cancer (N=175), African American Cancer Epidemiology Study (N=155), and the North Carolina Ovarian Cancer Study (N=136). Protein imaging data was collected using AKOYA Biosciences OPALTM IHC Kit with image analysis completed using Vectra®3 Automated Quantitative Pathology Imaging System. Spatial trajectories using G statistic were computed for samples with at least 8 positive cells for a cell type. FPCA was applied to the spatial curves with the top two components (FPC1, FPC2) associated with survival, adjusting for stage, age of diagnosis, and abundance of the cell population (high vs low using 1% threshold). A second model was fit to assess interaction between the abundance and spatial clustering. Analyses were completed for each study with results combined using a random-effect meta-analysis. Results: From the model without spatial information, we observed that high abundance of CD3+ (hazard ratio (HR): 0.81, 95% confidence interval (0.66, 0.98)) and CD3+CD8+ cells (HR: 0.64 (0.52, 0.79)) were associated with improved survival. The model with both abundance and spatial clustering detected a significant effect for CD3+CD8+ clustering (FPC1 HR: 1.17 (1.04, 1.33)) and a borderline association for CD3+ cells (FPC1 HR: 1.06 (0.99, 1.14)). When fitting a model with interactions for abundance and spatial clustering, a significant interaction for CD3+ cells (HR: 1.23 (1.07, 1.42)) and a borderline interaction for CD3+CD8+ cells (HR: 1.19 (0.98, 1.43)) was observed. Hence, we estimated the HRs for 4 tumor types (high/low abundance and high/low spatial clustering). We observed that patients with high abundance but low spatial clustering of CD3+ and CD3+CD8+ cells had the improved survival, with HRs for the high abundance / low spatial clustering group being 0.74 and 0.41, respectively. Discussion: In studying the HGSOC TME using spatial proteomics and FPCA, we found that not only is the abundance of T cell populations related to survival, but also the spatial clustering of these cell populations, with improved survival for women with tumors with diffuse T cell infiltration. Citation Format: Brooke L. Fridley, Alex C. Soupir, Daisy Liao, Chase Sakitis, Joellen Schildkraut, Andrew B. Lawson, Mary K. Townsend, Shelley Tworoger, Kathryn L. Terry, Julia Wrobel, Lauren Cole Peres. Functional data analysis of spatial protein imaging data using spatial trajectories with application to ovarian cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 6844.

Abstract Background: Tumor-infiltrating lymphocytes are associated with improved ovarian cancer (OC) survival in White women but not Black women despite similar T cell abundance. We evaluated whether exhaustion or localization of T cell infiltration may attenuate the survival advantage among Black women using a spatially informed compartment-specific approach to account for heterogeneity of whole tissue sections (WTS). Methods: Among 221 Black OC cases in the African American Cancer Epidemiology Study, multiplex immunofluorescence and HALO image analysis quantified total (CD3+) and cytotoxic (CD3+CD8+) T cells, along with exhausted (TIM3+) and terminally exhausted (TIM3+PD1+) subsets in FFPE WTS. PANCK-defined tumor and stroma were refined by excluding mixed interface regions detected using point-pattern intensities. Logit-transformed proportions were compared with paired t-tests. Cox proportional hazards estimated associations of compartment features with overall survival, adjusting for clinical factors. Results: Black OC cases were primarily advanced stage (58%) and high-grade serous carcinoma (HGSC; 69%); mean follow-up was 6.4 years. In tumor, 60% were T cell infiltrated (≥2% of tumor cells), and the median proportion of cytotoxic cells was 30%. Stroma contained 50% more T cells (p=9.1x10-12) but lower cytotoxic fractions (ΔT-S = -8%; p=3.9x10-19). Immune-excluded tumors (<2% tumor T cells) displayed strong stromal enrichment of total and cytotoxic T cells. Exhausted T cells were rare (1%) with tumors showing higher exhaustion (ΔT-S=1%, p=1.6x10-3) and terminal fractions (ΔT-S=15%, p=8.8x10-18) than stroma. Neither tumor nor stroma T cell abundance was associated with survival in univariate models. However, in compartment-adjusted models, tumoral cytotoxic T cells were associated with improved survival (hazard ratio [HR]=0.72, 95% confidence interval [CI]=0.53, 0.97), whereas stromal cytotoxic T cells were associated with worse survival (HR=1.65, CI=1.14, 2.37). A larger tumor-stroma gradient in cytotoxic T cells (> tumor) was similarly protective (HR=0.65, CI=0.48, 0.89). In HGSC, that had higher infiltration and exhaustion, infiltrated tumors had improved survival with higher exhaustion (HR=0.83, CI=0.70, 0.98) and terminal exhaustion (HR=0.88, CI=0.79, 0.99), while immune-excluded tumors did not. Conclusion: Distinct compartment-specific T cell and exhaustion patterns with divergent survival implications were observed in Black women with OC. Cytotoxic T cells confer a survival benefit only when tumor-dominant rather than stroma-dominant, highlighting the importance of characterizing an immune-exclusion phenotype. Improved survival with tumor T cell exhaustion likely indicates active antitumor immunity. WTS-based spatial profiling is essential for accurate interpretation of immune-survival relationships in Black women with OC. Citation Format: Brett M. Reid, Alex C. Soupir, Anthony J. Alberg, Elisa V. Bandera, Melissa L. Bondy, Michele L. Cote, Kristin Haller, Theresa Hastert, Carlos Moran Segura, Jonathan V. Nguyen, Edward S. Peters, Paul D. Terry, Andrew B. Lawson, Jeffrey R. Marks, Brooke L. Fridley, Joellen M. Schildkraut, Lauren Cole Peres. Tumor-stroma spatial context of T cell infiltration and exhaustion as determinants of ovarian cancer survival in Black women [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 6814.

Spatial proteomic imaging technologies enable the simultaneous assessment of immune cell abundance and spatial organization within the tumor microenvironment. Spatial clustering is commonly summarized using measures such as Ripley's K or nearest neighbor G-functions at a fixed radius. However, these approaches depend on scale selection and may obscure biologically relevant patterns occurring across spatial ranges. We propose a functional data analysis (FDA) framework to model spatial clustering trajectories derived across a continuum of radii. Functional principal component analysis (FPCA) was used to summarize dominant modes of spatial variation, and resulting scores were incorporated into Cox proportional hazards models as both main effects and interaction with immune cell abundance. The approach was applied to multiplex immunofluorescence data from five ovarian cancer studies, comprising 773 high grade ovarian serous tumors. Analyses focused on CD3+ and CD8+ T cell populations within the tumor compartment of the tissue, adjusting for age at diagnosis and cancer stage, with study-specific estimates combined using random-effects meta-analysis. Higher abundance of both T cells and CD8+ T cells was consistently associated with improved overall survival. Beyond abundance, spatial features captured by the leading functional principal component were independently associated with survival, particularly for CD8+ T cells. Interaction models further showed that the prognostic effect of immune infiltration depended on spatial clustering, with tumors characterized by high abundance and low spatial clustering exhibiting the most favorable outcomes. These findings indicate that spatial organization provides complementary prognostic information beyond abundance alone and suggests that more diffuse immune infiltration may reflect more effective anti-tumor activity in ovarian cancer. Overall, FDA offers a flexible and interpretable framework for modeling spatial clustering across scales and identifying prognostic spatial features not captured by fixed-radius or distance analyses.

Abstract Background: Black women experience poor survival from epithelial ovarian cancer (EOC), for all stages at diagnosis and EOC histotypes. We present findings of the contribution of African ancestry and an ancestry-related genotype in the Atypical Chemokine Receptor 1 (ACKR1) gene to EOC survival among a cohort of Black/African American women. We also report associations with tumor molecular features that may explain the potential underlying biology related to the Duffy-null ACKR1 genotypes. Methods: The relationship between global African ancestry and the rs2814778 SNP in the promotor region of the ACKR1 gene and survival was determined in a cohort of 408 Black women with EOC (275 with high grade serous ovarian cancer (HGSC)) who participated in the population-based African American Cancer Epidemiology Study (AACES) using a Bayesian modeling approach adjusting for stage, age at diagnosis, the Yost index for socioeconomic status, education, and ovarian cancer family history. Proportion of global African ancestry and the ACKR1 genotype (Duffy-null (CC) vs. TC/ TT) were determined from germline DNA. Tumor molecular features including gene expression using RNAseq, tumor immunity (i.e., T-cell abundance) measured using multiplex immunofluorescence (mIF), and homologous recombination deficiency (HRD) derived from whole exome sequencing data were generated from HGSC tumors. Results: The Duffy-null genotype was present in 70.6% EOC cases overall, and 72.0% of HGSC. The prevalence was higher among individuals with high African ancestry (African ancestry >83.6%) than those with lower African ancestry (83.2% and 58.3%, respectively). The Duffy-null genotype was associated with improved survival among EOC cases (adjusted Hazard Ratio (HR)=0.59, 95% credible intervals (CI): 0.35-0.96). The corresponding HR for HGSC was 0.61 (95% CI: 0.34-1.12). However, global African ancestry—as indicated by a 1% increase in percent African ancestry on the logit scale—was suggestively associated with worse survival, especially for HGSC, with an HR of 1.31 (95% CI: 0.89-1.90). Additionally, we examined the relationship between the CC vs. TC/ TT genotypes and tumor molecular features in HGSC. The Duffy-null genotype was associated with lower ACKR1 expression, lower cytotoxic T cell abundance, and higher HRD in HGSC. Among EOC subjects, mIF-derived myeloid cells (CD11b+) showed lower abundance in those with high African ancestry, which is consistent with worse survival. Conclusion: The Duffy-null genotype is associated with ∼40% decreased EOC mortality in Black women. Lower T-cell infiltrates may reflect lower immune surveillance and may coincide with higher HRD in HGSC among Duffy-null individuals. The HRD finding, in particular, may explain the improved prognosis observed with the Duffy-null genotype, as our group has previously shown that HRD status was associated with better survival in Black women with HGSC. Citation Format: Joellen M. Schildkraut, Jeffrey R. Marks, Xintian Song, Yao Xin, Anthony J. Alberg, Lauren C. Peres, Katherine Anne Lawson-Michod, Lindsay Jane Collin, Jennifer A. Doherty, Andrew B. Lawson, on behalf of the African American Cancer Epidemiology Study. African ancestry, Duffy-null genotype, and epithelial ovarian cancer in a cohort of Black women [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 2340.

Agricultural pesticide use represents one of the most geographically patterned environmental systems, yet most prior research has focused largely on individual compounds rather than correlated exposure regimes. We conducted a cross-sectional analysis using modeled pesticide application intensity and Alzheimer's Disease (AD) dementia prevalence at the county-level across the United States. Stability-based Elastic Net screening and clustering were used to identify exposure groupings, and associations with AD prevalence were estimated using adjusted regression models. Out of 462 total pesticides screened, 112 demonstrated high selection stability and were grouped into 25 exposure clusters. Twenty clusters were significantly associated with AD dementia prevalence (p<0.05). The strongest positive associations were observed for a soil fumigation/nematicide system, an herbicide-dominant vegetation control regime, and a neuroactive insecticide system. Neuroactive insecticides and soil-intensive treatment systems were disproportionately represented among positively associated clusters, whereas systems dominated by phenoxy- and photosystem II inhibiting herbicides were more frequently aligned with inverse gradients. The fully adjusted model explained 59% of between-county variance (baseline R2 = 0.44). Findings suggest that pesticide mixtures are associated with geographic heterogeneity in AD dementia prevalence and warrant higher-resolution, longitudinal investigation.

Background:Ovarian cancer is a leading cause of gynecologic cancer mortality, with Black women experiencing 5-year survival rates of only 41%. Disproportionate air pollution exposure may impact survival. We evaluated associations of fine particulate matter (PM2.5) and nitrogen dioxide (NO2) exposure with survival among Black women with epithelial ovarian cancer using data from the California Cancer Registry (CCR, n = 540) and the multi-state African American Cancer Epidemiology Study (AACES, n = 766). Methods:Annual PM2.5 and NO2 levels were estimated at a 1 km resolution using well-validated ensemble-based prediction models derived from the Socioeconomic Data and Application Center and assigned to the participants' residential addresses per their year of diagnosis (2004-2016). Weibull accelerated failure time models with participant-level frailty were used to assess air pollutant exposure associations with overall survival. Results:Average PM2.5 and NO2 exposures were 11.3 μg/m³ and 25.8 ppb in CCR and 9.7 μg/m³ and 17.5 ppb in AACES. There was little evidence of an association between air pollution exposures and survival, with event time ratios (> 1 indicate longer survival) in CCR of 1.08 (95% CI = 0.97, 1.20) per 1 μg/m³ PM2.5 and 1.07 (95% CI = 0.99, 1.15) per 10 ppb NO2, and in AACES of 1.00 (95% CI = 0.93, 1.07) per 1 μg/m³ PM2.5 and 1.04 (95% CI = 0.91, 1.19) per 10 ppb NO2. Conclusions:Findings were modest and consistent across both cohorts and sensitivity analyses, supported by the use of advanced exposure modeling. Future research should use time-varying, long-term exposure data and examine interactions with occupation, physical activity, and neighborhood stressors.

The ability to manage ill health and care needs might be affected by who a person lives with. This study examined how the risk of unplanned hospitalisation and transition to living in a care home varied according to household size and co-resident multimorbidity. Here we show results from a cohort study using Welsh nationwide linked healthcare and census data, that employed multilevel multistate models to account for the competing risk of death and clustering within households. The highest rates of unplanned hospitalisation and care home transition were in those living alone. Event rates were lower in all shared households and lowest when co-residents did not have multimorbidity. These differences were more substantial for care home transition. Therefore, living alone or with co-residents with multimorbidity poses additional risk for unplanned hospitalisation and care home transition beyond an individual's sociodemographic and health characteristics. Understanding the mechanisms behind these associations is necessary to inform targeted intervention strategies.

In Bayesian disease mapping, defining the neighborhood structure is crucial when fitting the conditional auto-regressive model. Yet, there has been little assessment of how different structures affect the model performance in case of fine-scale data. This paper explores this gap. In a case study examining COVID-19 pandemic effects, 2020 mortality is contrasted with pre-pandemic rates in small areas in Limburg (Belgium). Data are modeled using BYM and BYM2, with three broadening queen-neighborhood structures up to the fifth-order neighbors and two weight schemes. A simulation study assesses model performance in reproducing the pairwise spatial correlation at different neighbor orders. Models are compared regarding WAIC, goodness-of-fit, parameter estimates, and computation time. Results show that the order-based weight matrix performs better than the binary matrix. The simple first-order neighborhood structure shows comparable performance to larger higher-order structures while requiring much less computation time. The BYM model is more impacted by the choice of the neighborhood as compared to the BYM2 model. Our findings suggest minimal advantages in employing higher-order neighborhood matrices. In conclusion, our study indicates that opting for a simple first-order neighborhood structure is a pragmatic and suitable choice when applying a conditional auto-regressive model to fine-scale data in Bayesian disease mapping.

Ovarian cancer (OC) is the fifth-leading cause of cancer mortality among women in the US. Black women experience significantly lower OC survival than White women. Evidence suggests that this disparity is not solely the result of barriers to healthcare access but may also be impacted by other factors, including neighborhood social characteristics. To investigate this further, this study developed an approach for remotely estimating the degree of physical disorder (PD) in neighborhoods using structured audits of Google Street View imagery for participants in the AACES, a multi-site population-based study of Black women newly diagnosed with OC. We then assessed whether neighborhood disadvantage (ND) and PD were associated with overall survival. We fit Weibull accelerated failure time models to assess the association of both PD and ND with survival among 471 Black women with OC (n=317 deaths). Both PD (Event Time Ratio (ETR): 0.99, 95% CI: 0.98, 1.00) and Area Deprivation Index (ETR: 0.96, 95% CI: 0.94, 1.00) were associated with shorter survival. The results suggest that both physical and social neighborhood characteristics may impact survival in woman with OC, but further research is warranted.

BACKGROUND:Numerous studies have documented the negative impact of cigarette smoking on ovarian cancer survival, but the participants in these prior studies were predominantly White women. In comparison, Black women experience significantly worse ovarian cancer survival, which may be due in part to dissimilar risk factor profiles or factors associated with survival. We therefore examined the association between cigarette smoking and survival in a cohort of Black women with ovarian cancer. METHODS:This study included participants in the multi-site population-based African American Cancer Epidemiology Study (AACES), a prospective cohort study of 592 Black women with epithelial ovarian cancer followed up for an average of 5.5 years. Cox proportional hazards models were fit to estimate the association between cigarette smoking status (current and former smoking vs. never smoking) and all-cause mortality adjusting for sociodemographic, lifestyle, and clinical factors. RESULTS:Compared with women who never smoked cigarettes, women who currently smoked cigarettes experienced worse, but not statistically significant, survival (HR 1.41; 95 % CI 0.95- 2.10), whereas women who had quit smoking had comparable survival (HR 1.06; 95 %CI 0.82-1.35). Among former smokers, the association among those who quit smoking within the past five years was of similar magnitude as for current smoking (HR 1.37; 95 % CI 0.97-1.94) but no risk was observed among those who quit for > 5 years. CONCLUSION:Black women with epithelial ovarian cancer who were current smokers experienced worse survival than those who never smoked cigarettes. Even though this association was not statistically significant, the magnitude of the association is similar to prior studies comprised predominantly of White women. Ensuring access to evidence-based smoking cessation strategies represents a potential avenue for reducing mortality in Black women with ovarian cancer.

Objective:To examine how the risk of unplanned admission to hospital and transitioning to live in a care home by number of long term conditions varies by household size. Design:Retrospective cohort study. Setting:Wales Census 2011 household data, linked to the Welsh Secure Anonymised Information Linkage (SAIL) Databank, 27 March 2011 to 26 March 2016. Participants:391 686 residents of Wales recorded in the Wales Census on 27 March 2011, aged ≥65 years, living in Welsh households of one to six residents, registered with a general practitioner contributing data to the SAIL Databank. Main outcome measures:Time to the first unplanned hospital admission and time to transition from living at home in the community to living in a care home, for individuals with 0-1, 2-3, or ≥4 long term conditions living alone or in households with two residents or three or more residents. Results:Of the 391 686 individuals included, 36.8% lived alone, 54.0% lived in households of two, and 9.2% lived in households with three or more people. The number of long term conditions was strongly associated with the risk of hospital admission and transition to a care home. In those living in two person households, participants with ≥4 long term conditions versus those with 0-1 long term conditions had a higher risk of unplanned hospital admissions (adjusted hazard ratio 2.51, 95% confidence interval (CI) 2.47 to 2.55; crude event rate 180.1 (95% CI 178.5 to 181.7) v 54.8 (53.9 to 55.7) per 1000 person years) and of transitioning to live in a care home (adjusted hazard ratio 2.57, 2.49 to 2.66; crude event rate 7.2 (6.9 to 7.5) v 1.40 (1.3 to 1.5) per 1000 person years). Household size was associated with an increased risk of both outcomes but more strongly with transition to a care home than unplanned hospital admission. The risk of unplanned hospital admissions was higher for people with 0-1 long term conditions who lived alone than for those who lived in a two person household (adjusted hazard ratio 1.19, 95% CI 1.17 to 1.22; crude event rate 74.9 (95% CI 73.4 to 76.4) v 54.8 (53.9 to 55.7) per 1000 person years) and for transitioning to live in a care home (adjusted hazard ratio 1.48, 1.42 to 1.54; crude event rate 5.4 (5.0 to 5.8) v 1.4 (1.3 to 1.5) per 1000 person years). The association between the number of long term conditions and both outcomes varied by household size. Individuals with 0-1 long term conditions and living alone showed a higher risk of transitioning to live in a care home than individuals with 2-3 long term conditions living in two person households. Conclusions:In this study, the number of long term conditions was strongly associated with the risk of hospital admission and transition to living in a care home, and this association was less pronounced among those living alone. The risk of transitioning to live in a care home was higher for people with 0-1 long term conditions who lived alone than for those with 2-3 long term conditions who lived in two person households. These findings emphasise the need for personalised strategies that reduce the risk of unplanned admissions to hospital and support independent living, and that consider both the degree of multimorbidity and household size.

Focusing on data commonly found in public health databases and clinical settings, Bayesian Disease Mapping: Hierarchical Modeling in Spatial Epidemiology provides an overview of the main areas of Bayesian hierarchical modeling and its application to the geographical analysis of disease. The book explores a range of topics in Bayesian inference and

Abstract Introduction: Black women with epithelial ovarian cancer (EOC) have worse survival compared to other racial groups, and the causes of these poor outcomes remain unclear. Compared to other racial groups, Black women are disproportionately affected by comorbid conditions which can adversely impact cancer care and outcomes. Thus, we examined the association of pre-diagnostic comorbid conditions and their associated medications with survival among Black women with EOC. Methods: Using data from Black women with EOC in the African American Cancer Epidemiology Study, we evaluated the self-reported Charlson comorbidity index (CCI) and three cardiometabolic comorbidities (diabetes mellitus, hypertension, and hyperlipidemia). We also characterized whether women with each cardiometabolic condition were using medication for their condition. Kaplan-Meier survival curves and log-rank tests were used to examine survival by the CCI, each cardiometabolic comorbidity, and medication use. Cox proportional hazards regression models were used to examine the association of comorbid conditions and medications with survival while adjusting for age at diagnosis, stage, histotype, and study site. Results: Among 592 Black women with EOC, 35% had a CCI ≥2, and the prevalence of diabetes, hyperlipidemia, and hypertension was 19%, 31%, and 62%, respectively. Among women with each cardiometabolic condition, the prevalence of medication use for diabetes, hypertension, and hyperlipidemia was 73%, 78%, and 60%. In bivariate analyses, women with a higher CCI, diabetes, hyperlipidemia, and hypertension had worse survival compared to women without these conditions (P<0.05). However, when adjusting for prognostic factors, only a high CCI (≥2 vs. 0) and diabetes were significantly associated with higher hazard of death (HR=1.36, 95% CI=1.06-1.74 and HR=1.40, 95% CI=1.08-1.81, respectively). Investigating the independent associations of each condition included in the CCI with survival revealed that diabetes was largely driving the association of CCI with survival. When considering medication use, women with diabetes, irrespective of medication use, women with hypertension not on medication, and women with hyperlipidemia on medication had worse survival compared to women without these conditions in bivariate models (P<0.05). Compared to women without diabetes, women with diabetes on medication and women with diabetes not on medication had a statistically significant higher hazard of death in multivariable models (HR=1.39, 95% CI=1.02-1.88 and HR=1.63, 95% CI=1.04-2.57, respectively). No associations with survival were observed when considering medication use for hypertension and hyperlipidemia after adjusting for prognostic factors. Conclusion: Similar to prior studies among White women with EOC, diabetes, regardless of medication status, was strongly associated with poorer survival among Black women with EOC. Citation Format: Alicia Richards, Courtney E. Johnson, Anthony J. Alberg, Elisa V. Bandera, Melissa Bondy, Michele L. Cote, Theresa A. Hastert, Kristen Haller, Jeffrey R. Marks, Edward S. Peters, Paul D. Terry, Andrew B. Lawson, Joellen M. Schildkraut, Lauren C. Peres. Pre-diagnostic comorbid conditions and survival among black women with ovarian cancer [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 4829.

BackgroundMediation by multiple agents can affect the relation between neighborhood deprivation and segregation indices and ovarian cancer survival. In this paper, we examine a variety of potential clinical mediators in the association between deprivation indices (DIs) and segregation indices (SIs) with all-cause survival among women with ovarian cancer in the African American Cancer Epidemiology Study (AACES).MethodsWe use novel Bayesian multiple mediation structural models to assess the joint role of mediators (stage at diagnosis, histology, diagnostic delay) combined with the DIs and SIs (Yost, ADI, Kolak's URB, ICE-income) and a set of confounders with survival. The confounder set is selected in a preliminary step, and each DI or SI is included in separate model fits.ResultsWhen multiple mediators are included, the total impact of DIs and SIs on survival is much reduced. Unlike the single mediator examples previously reported, the Yost, ADI and ICE-income indices do not display significant direct effects. This suggests that when important clinical mediators are included, the impact of neighborhood SES indices is significantly attenuated. It is also clear that certain behavioral and demographic measures such as physical activity, smoking, or adjusted family income do not have a significant role in survival when mediated by clinical factors.ConclusionMultiple mediation via clinical and diagnostic-related measures reduces the contextual effects of neighborhood measures on ovarian cancer survival. The robust association of the Kolak URB index on survival may be due to its relevance to access to care, unlike SES-based indices whose impact was significantly reduced when important clinical mediators were included.

Dengue fever poses a significant public health burden in tropical regions, including Thailand, where periodic epidemics strain healthcare resources. Effective disease surveillance is essential for timely intervention and resource allocation. Various methods exist for spatiotemporal cluster detection, but their comparative performance remains unclear. This study compared spatiotemporal cluster detection methods using simulated and real dengue surveillance data from Thailand. A simulation study explored diverse disease scenarios, characterized by varying magnitudes and spatial-temporal patterns, while real data analysis utilized monthly national dengue surveillance data from 2018 to 2020. Evaluation metrics included accuracy, sensitivity, specificity, positive predictive value, and negative predictive value. Bayesian models and FlexScan emerged as top performers, demonstrating superior accuracy and sensitivity. Traditional methods such as Getis Ord and Moran's I showed poorer performance, while other scanning-based approaches like spatial SaTScan exhibited limitations in positive predictive value and tended to identify large clusters due to the inflexibility of its scanning window shape. Bayesian modeling with a space-time interaction term outperformed testing-based cluster detection methods, emphasizing the importance of incorporating spatiotemporal components. Our study highlights the superior performance of Bayesian models and FlexScan in spatiotemporal cluster detection for dengue surveillance. These findings offer valuable guidance for policymakers and public health authorities in refining disease surveillance strategies and resource allocation. Moreover, the insights gained from this research could be valuable for other diseases sharing similar characteristics and settings, broadening the applicability of our findings beyond dengue surveillance.

Abstract Introduction: The spatial contexture of the tumor immune microenvironment (TIME) has been associated with survival among cancer patients, including women with high-grade serous ovarian cancer (HGSOC). However, an assessment of statistical approaches to quantify the spatial characteristics of the TIME has not been conducted. Moreover, it is unknown which statistical approach is most sensitive in identifying spatial co-localization of two cell types in the TIME. Methods: Using the R package scSpatialSIM that simulates spatial single-cell protein data (i.e., multiplex immunofluorescence data), we completed a simulation study to compare four methods for assessing co-localization of cell types in the TIME (Ripley’s K, Nearest Neighbor G, Dixon’s segregation statistic, and a spatial interaction variable). The simulation study varied the cell abundance (e.g., low vs high), the co-localization (e.g., co-localization, random, segregation), and the variation in the marked point process. In addition to the simulated data, we assessed the four methods in large epidemiological studies (Nurses’ Health Study [NHSI/NHSII], New England Case Control Study [NECC], African American Cancer Epidemiology Study [AACES], and North Carolina Ovarian Cancer Study [NCOCS]) including 756 women with HGSOC. Multiplex immunofluorescence was conducted using AKOYA Biosciences OPALTM platform to measure cytotoxic T cells (CTLS) and regulatory T cells (Tregs). Cox proportional hazards regression models were fit to estimate the association of spatial co-localization of CTLS and Tregs with overall survival. Results: In the simulation study, we found that Ripley’s K was the most powerful approach for assessing co-localization for a variety of simulation scenarios. Surprisingly, the other approaches (Nearest Neighbor G, Dixon’s Segregation statistic, and spatial interaction statistic) were unable to detect the simulated co-localization of two cell types in most simulation scenarios. In the analysis of women with HGSOC in the epidemiological cohorts, we were able to detect significant co-localization of CTLS and Tregs in 40.2%, 12.2%, 21.2%, and 17.0% of samples using Ripley’s K, Nearest Neighbor G, Dixon’s statistic, and the interaction variable methods, respectively. However, the level of co-localization was not associated with survival. Conclusions: We found that Ripley’s K was the most powerful at detecting simulated co-localization of two cell types in the TIME. Similarly, Ripley’s K identified the highest level of significant spatial co-localization of CTLS and Tregs in the application to HGSOC. However, we did not observe an association of co-localization of CTLS and Tregs with overall survival. Future work is needed to develop powerful approaches for quantifying the spatial contexture of the TIME, along with assessment of the statistical properties of these methods. Citation Format: Brooke L. Fridley, Alex Soupir, Julia Wrobel, Christelle Colin-Leitzinger, Mary K. Townsend, Andrew B. Lawson, Kathryn L. Terry, Joellen M. Schildkraut, Shelley S. Tworoger, Lauren C. Peres. Comparison of spatial co-localization measures for studying the tumor immune microenvironment with application to ovarian cancer [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 3562.

Identification of areas of high disease risk has been one of the top goals for infectious disease public health surveillance. Accurate prediction of these regions leads to effective resource allocation and faster intervention. This paper proposes a novel prediction surveillance metric based on a Bayesian spatio-temporal model for infectious disease outbreaks. Exceedance probability, which has been commonly used for cluster detection in statistical epidemiology, was extended to predict areas of high risk. The proposed metric consists of three components: the area's risk profile, temporal risk trend, and spatial neighborhood influence. We also introduce a weighting scheme to balance these three components, which accommodates the characteristics of the infectious disease outbreak, spatial properties, and disease trends. Thorough simulation studies were conducted to identify the optimal weighting scheme and evaluate the performance of the proposed prediction surveillance metric. Results indicate that the area's own risk and the neighborhood influence play an important role in making a highly sensitive metric, and the risk trend term is important for the specificity and accuracy of prediction. The proposed prediction metric was applied to the COVID-19 case data of South Carolina from March 12, 2020, and the subsequent 30 weeks of data.

[This corrects the article DOI: 10.1371/journal.pone.0278515.].

This workshop provides a comprehensive introduction to advanced techniques for analyzing spatial and spatio-temporal data. While the examples used will be primarily from the health sciences, this four-day hands-on workshop is designed to equip PhD students, professors, and professional researchers with the skills to conduct cutting-edge research in various fields, including Geography, Epidemiology, Public Health, Biostatistics, Ecology, Sociology, and Political Science.