Sage Kim's research while affiliated with University of Illinois at Chicago and other places

1584 Background: In the US, disparities in lung cancer mortality exist for African American, Hispanic, and other minorities. Standard of care low dose CT screening (LDCT) detects early-stage disease and improves mortality, yet disparities are perpetuated in screening by eligibility criteria derived from cohorts underrepresenting these minorities. One such cohort is the National Lung Screening Trial (NLST) cohort which is 92% White. Consequently, guidelines for lung cancer screening may be insufficient to address the unique needs of diverse populations. We hypothesize that Artificial Intelligence prediction of future lung cancer risk from an individual’s LDCT can partially mitigate racial and ethnic disparities and improve health system practice guidelines by individualizing screening risk as compared to current general guidelines. Here, we benchmark a Resnet18 3D neural network trained on NLST LDCT images, Sybil, on the diverse patient population of the University of Illinois Health system (UIH) which is 20% White and 60% African American. Methods: A real-world cohort from UIH consisting of 1,450 CT studies was evaluated alongside 60,378 CT studies from the NLST cohort. All CT studies evaluated by the model were not used in model training. Using Youden’s J index as a probability cutoff, sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were evaluated. Receiver operating characteristic (ROC) and precision-recall (PR) curves were generated to assess model performance between cohorts. NLST data were truncated to achieve equivalent incidence of lung cancer with UIH when generated PR curves. Results: Multi-year prediction performance (ROC-AUC and PR-AUC) between cohorts are summarized (Table). For prediction of lung cancer within 1-year of LDCT in the UIH cohort, the model respectively demonstrated sensitivity, specificity, positive predictive value, and negative predictive value among White (0.80, 0.77, 0.54, 0.92), African American (0.84, 0.78, 0.39, 0.97) races and Hispanic (0.75, 0.73, 0.60, 0.84) and non-Hispanic (0.87, 0.77, 0.42, 0.97) ethnicities. Conclusions: Model performance was similar between the NLST (92% White) and a diverse, real-world cohort at UIH (20% White) though decreases in ROC-AUC performance in year 1 predictions and may be due to insufficient representation of minority populations during model training. Prospective studies involving larger and more representative patient populations should be conducted to further optimize the model and evaluate its clinical utility to improve lung cancer health equity in minority populations.[Table: see text]

Although many have investigated the impacts of minimum wage on a broad array of health outcomes, innovative policies surrounding broader employment policies have largely not been studied. To that end, this paper contributes in three ways. First, it discusses the rise in precarious employment. Then, it turns to the current federal framework of employment policies, namely minimum wage. Finally, it explores what a broader definition of employment policies could include and how future studies could use state, county, and municipal policymaking in this space to investigate ways in which they might contribute to reducing food insecurity and in turn, improve health outcomes.

(1) Background: Sarcopenia, or low skeletal mass index (SMI), contributes to higher lung cancer mortality. The SMI at third lumbar vertebrae (L3) is the reference standard for body composition analysis. However, there is a need to explore the validity of alternative landmarks in this population. We compared the agreement of sarcopenia identification at the first lumbar (L1) and second lumbar (L2) to L3 in non-Hispanic Black (NHB) and White (NHW) individuals with lung cancer. (2) Methods: This retrospective, cross-sectional study included 214 NHB and NHW adults with lung cancer. CT scans were analyzed to calculate the SMI at L1, L2, and L3. T-tests, chi-square, Pearson’s correlation, Cohen’s kappa, sensitivity, and specificity analysis were used. (3) Results: Subjects presented with a mean age of 68.4 ± 9.9 years and BMI of 26.3 ± 6.0 kg/m2. Sarcopenia prevalence varied from 19.6% at L1 to 39.7% at L3. Cohen’s kappa coefficient was 0.46 for L1 and 0.64 for L2, indicating weak and moderate agreement for the identification of sarcopenia compared to L3. (4) Conclusions: Sarcopenia prevalence varied greatly depending on the vertebral landmark used for assessment. Using L2 or L1 alone resulted in a 16.8% and 23.8% misclassification of sarcopenia in this cohort of individuals with lung cancer.

Background: Sarcopenia and inflammation are independent prognostic indicators of poor survival in lung cancer. Systemic inflammation contributes to sarcopenia or low muscle mass through increased catabolism and decreased synthesis of muscle protein. In turn, sarcopenia leads to inflammation in the muscle and exacerbation of systemic inflammation. Specific Aim: The aim of this study was to determine the association between sarcopenia and systemic biomarkers of inflammation including neutrophil to lymphocyte ratio (NLR), platelet to lymphocyte ratio (PLR), and albumin in diverse individuals with lung cancer and determine their use as prognostic indicators of sarcopenia. Methods: This retrospective cross-sectional study included 218 non-Hispanic Black (n=111) and White (n=107) individuals with lung cancer. Computed tomography scans were analyzed to calculate skeletal mass index (SMI) at the third lumbar vertebra. Sarcopenia was defined using published sex-specific criteria for SMI in healthy individuals then categorized into the binary variable of sarcopenia/no sarcopenia. NLR, PLR, and albumin were analyzed as both continuous and categorical variables, with NLR> 3, PLR >150, and albumin <3.5 defining moderate to high inflammation. T-tests, Mann Whitney U, Chi square, and adjusted Logistic Regression were used as appropriate. Results: Subjects presented with a mean age of 68.5 + 9.8 years, BMI of 26.3 + 5.9 kg/m2, 50.9% non-Hispanic Black, and 49.5% female with significantly higher prevalence of late-stage lung cancer (61.9% versus 38.1% early-stage, p<0.001) and no differences in age and BMI by race/ethnicity. Mean NLR (6.2 + 7.2), PLR (226.9 + 157.9), and albumin (3.2 + 0.6) indicated an inflammatory state in diverse individuals with lung cancer, with significantly higher NLR (8.5 + 9.7 versus 4.7 + 4.2, p<0.001) and PLR (268.9 + 204.5 versus 198.7 + 108.9, p<0.01) and significantly lower albumin (3.1 + 0.7 versus 3.4 + 0.5, p<0.01) in those with sarcopenia compared to those without sarcopenia. The odds of sarcopenia were 2.2 higher with NLR > 3 compared to NLR <3 (95% CI: 1.1, 4.4) and 2.5 higher with albumin <3.5 compared to albumin >3.5 (95% CI: 1.2, 5.3) after controlling for race/ethnicity, stage of cancer, age, and BMI. PLR > 150 was not a significant predictor of sarcopenia. Conclusion: Diverse individuals with sarcopenia of lung cancer present with higher inflammation compared to individuals with lung cancer but without sarcopenia as indicated by NLR, PLR, and albumin levels. Moderate to high levels of inflammation as indicated by high NLR and low albumin may help predict sarcopenia status in diverse individuals with lung cancer. Citation Format: Cecily A. Byrne, Giamila Fantuzzi, Vanessa M. Oddo, Sage Kim, Timothy J. Koh, Sandra L. Gomez. Inflammatory biomarkers by sarcopenia status in diverse individuals with lung 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 7649.

Background: Despite smoking less, Black men experience higher incidence of lung cancer compared to white men. This health disparity is even more prevalent in Black men from Chicago, suggesting local neighborhood factors impacting tumorigenesis. Preliminary data indicate that, compared to white men, Black men are more likely to reside in neighborhoods with high levels of violence and have elevated levels of hair cortisol. Objective: To understand the link between lung tumorigenesis and exposure to neighborhood violence, our objective is to investigate the chromatin recruitment of the receptor for cortisol, called the glucocorticoid receptor (GR), and to correlate this recruitment with target gene expression levels. Methods: Utilizing CUT & RUN, we identified the gene binding sites of GR using 15 lung tumor and corresponding healthy tissue samples from lung cancer patients living in Chicago. GR recruitment to chromatin was correlated with the neighborhood homicide rate obtained using patients’ zip codes. Spatial transcriptome profiles for these 15 lung tumor samples were obtained from the Space Ranger pipeline and gene expression hot spots of GR target genes were visualized using Loupe Browser (10x Genomics). Results: GR recruitment to regulatory regions of chromatin increases with greater zip code level violence rates. In patients from high violence neighborhoods, tumor samples had higher magnitude of binding compared to normal neighboring tissue. In tumor samples from patients living in high violence compared to low violence neighborhoods, GR binding is seen in genes that are associated with increased tumor aggressiveness. Hot-spot analysis of these genes found strong co-expression of genes associated with proliferation and invasion in samples from high violence but not low violence neighborhoods. Conclusion: Exposure to neighborhood violence may impact tumor biology via increased GR recruitment to genes associated with proliferation and invasion. Citation Format: Hannah Heath, Jin Young Yoo, Vani Sharma, Hannah McGee, Aiman Soliman, Abeer Mohamed, Alicia K. Matthews, Robert Winn, Zeynep Madak-Erdogan, Sage Kim. Exposure to neighborhood violence rewires glucocorticoid receptor binding and drives proliferation and invasiveness in lung tumor samples [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 6275.

Objectives The primary objective was to determine differences in Social Vulnerability Index (SVI) scores among minorities (African-Americans and Hispanics) with acute pancreatitis (AP) compared with non-Hispanic whites (NHWs) with AP. The secondary objectives were to determine differences in diet, sulfidogenic bacteria gene copy numbers (gcn) and hydrogen sulfide (H2S) levels between the 2 groups. Materials and Methods Patients with AP were enrolled during hospitalization (n = 54). Patient residential addresses were geocoded, and the Centers for Disease Control and Prevention’s SVI scores were appended. Dietary intake and serum H2S levels were determined. Microbial DNAs were isolated from stool, and gcn of sulfidogenic bacteria were determined. Results Minorities had higher SVI scores compared with NHWs ( P = 0.006). They also had lower consumption of beneficial nutrients such as omega-3 fatty acids [stearidonic ( P = 0.019), and eicosapentaenoic acid ( P = 0.042)], vitamin D ( P = 0.025), and protein from seafood ( P = 0.031). Lastly, minorities had higher pan-dissimilatory sulfite reductase A ( pan-dsrA ) gcn ( P = 0.033) but no significant differences in H2S levels ( P = 0.226). Conclusion Minorities with AP have higher SVI compared with NHWs with AP. Higher SVI scores, lower consumption of beneficial nutrients, and increased gcn of pan-dsrA in minorities with AP suggest that neighborhood vulnerability could be contributing to AP inequities.

The increasing rates of cancer incidence are disproportionately borne by populations that are ineligible for screening and historically marginalized populations. To address this need, our community-centered model seeks to catalyze the widespread diffusion of evidence-based information and resources (e.g., community-based organizations, federally qualified health centers) to reduce the risks of cancer, chronic disease, and other conditions. In this study, we tested whether improving personal health literacy (i.e., confidence in seeking information) and enabling successful information transfer (i.e., intention to share the specific information learned through the program) among community residents could contribute to greater diffusion intention (i.e., number of network members with whom residents plan to share information and resources). The current study used post-intervention surveys, which were administered to Chicago residents who were 18 years or older and had participated in the program. Among the 1499 diverse Chicago residents, improved personal health literacy was associated with greater diffusion intention (ORs = 2.00–2.68, 95% CI [1.27–4.39], p ≤ 0.003). Successful information transfer was associated with greater diffusion, especially for cancer and other chronic disease risk reductions (ORs = 3.43–3.73, 95% CI [1.95–6.68], p < 0.001). The findings highlight the potential gains for health equity through sustainable, scalable, multi-sectoral partnerships.

Health disparities persist among Black men, notably in the context of lung cancer and stress-related health outcomes. This study explores these disparities through a community-based participatory research (CBPR) approach, citizen science, and social network theory, leveraging the expertise and trust of Black barbers as community leaders. The purpose is to understand the nuanced connections between stress and lung cancer in this demographic. Engaging 161 Black men across four Chicago neighborhoods, the study successfully collected hair samples and survey data, emphasizing the importance of culturally sensitive recruitment strategies. Findings highlight the effectiveness of the collaboration, showcasing the role of barbershops as community hubs for research. The study concludes by advocating for sustained partnerships with community leaders, emphasizing transparency in research communication, and promoting culturally grounded approaches to address health disparities and enhance research participation among underrepresented populations.

Background: Since the onset of the COVID-19 pandemic, multiple public health interventions have been implemented to respond to the rapidly evolving pandemic and community needs. This article describes the scope, timing, and impact of coordinated strategies for COVID-19 vaccine uptake in Chicago for the first year of vaccine distribution. Methods: Using a series of interviews with public health officials and leaders of community-based organizations (CBOs) who participated in the implementation of the citywide COVID-19 vaccine outreach initiatives, we constructed a timeline of vaccine outreach initiatives. The timeline was matched to the vaccine uptake rates to explore the impact of the vaccine outreach initiatives by community area. Finally, we discussed the nature of policy initiatives and the level of vaccine uptake in relation to community characteristics. Results: The Chicago Department of Public Health (CDPH) implemented myriad vaccine outreach strategies, including mass vaccination sites, improved access, and community-level vaccine campaigns. Protect Chicago+ was the primary vaccine outreach effort initiated by the CDPH, which identified 15 highly vulnerable community areas. More than 2.7 million (67%) Chicagoans completed the vaccine regimen by December 2021. Black (51.3%) Chicagoans were considerably less likely to be vaccinated than Asian (77.6%), White (69.8%), and Hispanic (63.6%) Chicago residents. In addition, there were significant spatial differences in the rate of COVID-19 vaccine completion: predominantly White and Hispanic communities, compared with Black communities, had higher rates of vaccine completion. Conclusions: The community outreach efforts to improve COVID-19 vaccine uptake in Chicago have shown the importance of community-engaged approaches in pandemic responses. Despite citywide efforts to build community infrastructure, Black communities had relatively lower levels of vaccine uptake than other communities. Broader social restructuring to mitigate disinvestment and residential segregation and to ameliorate medical mistrust will be needed to prepare for future pandemics and disasters.

The study's goal was to look at providers' knowledge, attitudes, and behaviors regarding lung cancer screening among Black male smokers served by a federally qualified healthcare center. Participants in the study were interviewed in depth. Participants completed a short (5-10 minute) survey that assessed demographics, training, and attitudes toward lung cancer screening. For quantitative data, descriptive statistics were used, and for qualitative data, deductive thematic analysis was used. This study included ten healthcare professionals, the majority of whom identified as Black (80%) and were trained as advanced practice providers (60%). The majority of providers (90%) have heard of LDCT lung cancer screening; however, participants reported only being “somewhat” familiar with the LDCT eligibility criteria (70%). Despite generally positive attitudes toward LDCT, patient referral rates for screening were low. Barriers included a lack of provider knowledge about screening eligibility, a lack of use of shared decision-making tools, and patient concerns about screening risks. The reasons for the low referral rates varied, but they included a preference to refer patients for smoking cessation rather than screening, low screening completion and follow-up rates among referred patients, and a lower likelihood that Black smokers will meet pack-year requirements for screening. Additionally, providers discussed patient-level factors such as a lack of information, mistrust, and transportation. The study findings add to the body of knowledge about lung cancer knowledge and screening practices among providers in FQHC settings. This data can be used to create health promotion interventions aimed at smoking cessation and lung cancer screening in Black males and other high-risk smokers.