Vanderbilt University Medical Center Research Articles

Abstract PO2-12-02: Frailty among long-term breast cancer survivors

Abstract Background Cancer and its associated treatments may accelerate aging, resulting in increased frailty in survivors. However, the effect of breast cancer (BC) and treatment on frailty among long-term BC survivors has not been well investigated. This study aims to evaluate frailty prevalence and its association with clinical characteristics and treatment among 5-year BC survivors. Methods This study included 353 5-year BC survivors and 214 non-cancer control women aged ≥40 who were recruited to an ongoing breast health study at Vanderbilt University Medical Center between 2004 and 2017, and completed an online survey between May, 2022 and June 2023. Information collected includes demographics, lifestyle, disease history, and current health/functional status. A 26-item cumulative deficit index (CDI) was calculated as the proportion of deficits presented, including aging-related diseases, physical and mental health, activities of daily living, and social well-being, based on survey responses. CDI was categorized as robust (≤0.2), pre-frail (0.2-≤0.35), and frail ( >0.35). Cancer characteristics and treatment information was gathered from tumor registry. Multinomial logistic regression models were used to assess the association of CDI levels with case-control status, and BC clinical characteristics and cancer treatment among BC cases, with adjustment for age at survey (years), follow-up duration (5-≤10 and >10 years), and baseline Charlson comorbidity index (0, 1, and ≥2), and additional adjustment for BC stage and subtype in the analysis of treatment. Results The median time interval from BC diagnosis to survey was 12 years (range 5-34 years) for BC survivors. 90% of the participants were White. On average, BC patients were 3.8 years older at survey than controls [mean±SD (range) 64.5±9.2 (40-88) vs 60.7±10.3 (40-87), p< 0.001]. Among BC survivors, 41% had stage I, 39% stage II, and 8% stage III-IV cancer; 62% had a hormone receptor+/HER2- BC, followed by 22% HER2+ and 16% triple-negative BC. The majority of BC survivors underwent surgical resection (97%), and >50% received hormone therapy (59%), radiation therapy (53%), and chemotherapy (51%). BC survivors had a higher CDI score than controls (median 0.23 vs 0.19, p=0.01), with 34% and 18% of BC survivors being pre-frail and frail, compared with 25% and 16% in controls. After multivariable adjustment, BC survivors had a non-significantly increased prevalence of pre-frailty with odds ratio (OR) and 95% confidence interval (CI) of 1.49 (0.98-2.25), but a similar prevalence of frailty [OR (95% CI) = 1.08 (0.65-1.79)] compared to controls. BC survivors aged 40-65 had significantly elevated odds of pre-frailty [OR (95% CI) = 2.59 (1.44-4.63)]. Among BC survivors, frailty was significantly associated with stage [stage III-IV vs I: OR (95% CI) = 4.03 (1.27-12.78)] and chemotherapy [OR (95% CI) = 4.39 (1.71-11.30)], particularly among survivors aged ≥65 and 10-year survivors, but not BC subtype. Radiation therapy was associated with increased odds of pre-frailty among BC survivors aged 40-65. Conclusions This study suggests that long-term BC survivors younger than 65 are at increased odds of pre-frailty than non-cancer women. Frailty is associated with BC stage and chemotherapy among BC survivors, and radiotherapy among survivors younger than 65. More research is needed to investigate whether specific cancer treatment regimen is associated with frailty and other factors (e.g. lifestyle and genetics) that may modify the risk of frailty. The associations of frailty with cancer characteristics and treatments among breast cancer survivors Citation Format: Cong Wang, Jill De Vis, Kirsten Nguyen, Mason Alford, Brigitte Jia, Bapsi Chakravarthy, Xiao-Ou Shu. Frailty among long-term breast cancer survivors [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO2-12-02.

ADDRESSING PRIVACY AND ETHICAL CONSIDERATIONS IN HEALTH INFORMATION MANAGEMENT SYSTEMS (IMS)

This paper explores the crucial role of privacy and ethics in Health Information Management Systems (HIMS), addressing the inherent challenges and presenting innovative solutions through detailed case studies. We examine the significant privacy concerns, such as data breaches and unauthorized data sharing, alongside ethical issues like informed consent and equitable access to technology. Effective solutions including the Privacy-Preserving Distributed Analytics (PPDA) model and proactive bioethics committees, as demonstrated by institutions like Vanderbilt University Medical Center, illustrate successful strategies for managing these concerns. The paper emphasizes the importance of prioritizing privacy and ethics not merely as compliance requirements but as foundational elements essential to the trustworthiness and effectiveness of HIMS. It advocates for a continuous, proactive approach to address these issues as technology evolves and regulations change. Furthermore, we call for a collaborative effort among policymakers, healthcare providers, technologists, and patients to develop and refine HIMS that uphold the highest standards of privacy, ethics, and accessibility, thus enhancing the quality of care and health outcomes for all stakeholders.

A Real-time Prognostic Model for Venous Thromboembolic Events among Hospitalized Adults

BackgroundHospital-acquired venous thromboembolism (HA-VTE) is a leading cause of morbidity and mortality among hospitalized adults. Guidelines recommend use of a risk-prediction model to estimate HA-VTE risk for individual patients. Extant models do not perform well for broad patient populations and are not conducive to automation in clinical practice. ObjectiveTo develop an automated, real-time prognostic model for VTE during hospitalization among all adult inpatients using readily available data from the electronic health record (EHR). MethodsThe derivation cohort included inpatient hospitalizations (‘encounters’) for patients ≥16 years old at Vanderbilt University Medical Center between 2018-2020 (n=132,330). HA-VTE events were identified using ICD-10 codes. The prognostic model was developed using LASSO regression. Temporal external validation was performed in a validation cohort of encounters between 2021-2022 (n=62,546). Prediction performance was assessed by discrimination accuracy (C statistic) and calibration (integrated calibration index, ICI). ResultsThere were 1,187 HA-VTEs in the derivation cohort (9.0 per 1000 encounters) and 864 in the validation cohort (13.8 per 1000 encounters). The prognostic model included 25 variables, with placement of a central line among the most important predictors. Prediction performance of the model was excellent (C statistic: 0.891, 95% CI: 0.882-0.900; ICI: 0.001). The model performed similarly well across subgroups of patients defined by age, sex, race, and type of admission. ConclusionThis fully automated prognostic model uses readily available data from the EHR, exhibits superior prediction performance compared to existing models, and generates granular risk stratification in the form a predicted probability of HA-VTE for each patient.

Financial impact of integrated specialty pharmacy efforts to avoid oral anticancer medication waste.

The growing number of oral anticancer medications represents a significant portion of pharmacy spending and can be costly for patients. Patients taking oral anticancer medications may experience frequent treatment changes following necessary safety and effectiveness monitoring, often resulting in medication waste. Strategies to avoid medication waste could alleviate the financial burden of these costly therapies on the payer and the patient. To evaluate the impact on waste and cost avoidance of reviewing the amount of medication patients have on hand and the presence of upcoming follow-up (ie, provider visit, laboratory testing, or imaging) before requesting a prescription refill renewal for patients taking oral anticancer medications through an integrated health system specialty pharmacy. We performed a retrospective review of patients filling oral anticancer medications prescribed by a Vanderbilt University Medical Center provider and dispensed by Vanderbilt Specialty Pharmacy between January 1, 2020, and December 31, 2020. Specialty pharmacists received a system-generated refill renewal request for oral anticancer medications when the final prescription refill was dispensed, prompting the pharmacist to review the patient's medical record for continued therapy appropriateness and to request a new prescription. If the patient had a sufficient supply on hand to last until an upcoming follow-up (ie, provider visit, imaging, or laboratory assessment), the pharmacist postponed the renewal until after the scheduled follow-up. Patients were included in the analysis if the refill renewal request was postponed after review of the amount of medication on hand and the presence of an upcoming follow-up. Medication outcomes (ie, continued, dose changed, held, medication changed to a different oral anticancer medication, or discontinued) resulting from the follow-up were collected. Cost avoidance in US dollars was assigned based on the outcome of follow-up by calculating the price per unit times the number of units that would have been unused or in excess of what was needed if the medication had been dispensed before the scheduled follow-up. The average wholesale price minus 20% (AWP-20%) and wholesale acquisition cost (WAC) were used to report a range of costs avoided over 12 months. The total cost avoidance over 12 months associated with postponing refill renewal requests in a large academic health system with an integrated specialty pharmacy ranged from $549,187.03 using WAC pricing to $751,994.99 using AWP-20% pricing, with a median cost avoidance per fill of $366.04 (WAC) to $1,931.18 (AWP-20%). Refill renewal requests were postponed in 159 instances for 135 unique patients. After follow-up, medications were continued unchanged in only 2% of postponed renewals, 56% of follow-ups resulted in medication discontinuations, 32% in dose changes, 5% in medication changes, and 5% in medication holds. Integrated health system specialty pharmacist postponement of refill requests after review of the amount of medication on hand and upcoming follow-up proved effective in avoiding waste and unnecessary medication costs in patients treated with oral anticancer medications at a large academic health system.

Characteristics and Health Care Utilization of Patients With Housing Insecurity in the ED

Unstable housing and homelessness can exacerbate adverse health outcomes leading to increased risk of chronic disease, injury, and disability. However, emergency departments (EDs) have no universal method to identify those at risk of or currently experiencing homelessness. To describe the extent of housing insecurity among patients who seek care in an urban ED, including chief concerns, demographics, and patterns of health care utilization. This cross-sectional study included all adult patients presenting to the ED at Vanderbilt University Medical Center (VUMC), an urban tertiary care, level I trauma center in the Southeast US, from January 5 to May 16, 2023. The primary outcome was the proportion of ED visits at which patients screened positive for housing insecurity. Secondary outcomes included prevalence of insecurity by chief concerns, demographics, and patterns of health care utilization. Of all 23 795 VUMC ED visits with screenings for housing insecurity (12 465 visits among women [52%]; median age, 47 years [IQR, 32-48 years]), in 1185 (5%), patients screened positive for current homelessness or housing insecurity (660 unique patients); at 22 610 visits (95%), the screening result was negative. Of visits with positive results, the median age of patients was 46 years (IQR, 36-55 years) and 829 (70%) were among male patients. Suicide and intoxication were more common chief concerns among visits at which patients screened positive (132 [11%] and 118 [10%], respectively) than among those at which patients screened negative (220 [1%] and 335 [2%], respectively). Visits with positive results were more likely to be among patients who were uninsured (395 [33%] vs 2272 [10%]) and had multiple visits during the study period. A higher proportion of positive screening results occurred between 8 pm and 6 am. The social work team assessed patients at 919 visits (78%) with positive screening results. In this cross-sectional study of 23 795 ED visits, at 5% of visits, patients screened positive for housing insecurity and were more likely to present with a chief concern of suicide, to be uninsured, and to have multiple visits during the study period. This analysis provides a call for other institutions to introduce screening and create tailored care plans for patients experiencing housing insecurity to achieve equitable health care.

Phenomewide Association Study of Health Outcomes Associated With the Genetic Correlates of 25 Hydroxyvitamin D Concentration and Vitamin D Binding Protein Concentration.

While it is known that vitamin D deficiency is associated with adverse bone outcomes, it remains unclear whether low vitamin D status may increase the risk of a wider range of health outcomes. We had the opportunity to explore the association between common genetic variants associated with both 25 hydroxyvitamin D (25OHD) and the vitamin D binding protein (DBP, encoded by the GC gene) with a comprehensive range of health disorders and laboratory tests in a large academic medical center. We used summary statistics for 25OHD and DBP to generate polygenic scores (PGS) for 66,482 participants with primarily European ancestry and 13,285 participants with primarily African ancestry from the Vanderbilt University Medical Center Biobank (BioVU). We examined the predictive properties of PGS25OHD, and two scores related to DBP concentration with respect to 1322 health-related phenotypes and 315 laboratory-measured phenotypes from electronic health records. In those with European ancestry: (a) the PGS25OHD and PGSDBP scores, and individual SNPs rs4588 and rs7041 were associated with both 25OHD concentration and 1,25 dihydroxyvitamin D concentrations; (b) higher PGS25OHD was associated with decreased concentrations of triglycerides and cholesterol, and reduced risks of vitamin D deficiency, disorders of lipid metabolism, and diabetes. In general, the findings for the African ancestry group were consistent with findings from the European ancestry analyses. Our study confirms the utility of PGS and two key variants within the GC gene (rs4588 and rs7041) to predict the risk of vitamin D deficiency in clinical settings and highlights the shared biology between vitamin D-related genetic pathways a range of health outcomes.

The broad impact of cell death genes on the human disease phenome

Cell death mediated by genetically defined signaling pathways influences the health and dynamics of all tissues, however the tissue specificity of cell death pathways and the relationships between these pathways and human disease are not well understood. We analyzed the expression profiles of an array of 44 cell death genes involved in apoptosis, necroptosis, and pyroptosis cell death pathways across 49 human tissues from GTEx, to elucidate the landscape of cell death gene expression across human tissues, and the relationship between tissue-specific genetically determined expression and the human phenome. We uncovered unique cell death gene expression profiles across tissue types, suggesting there are physiologically distinct cell death programs in different tissues. Using summary statistics-based transcriptome wide association studies (TWAS) on human traits in the UK Biobank (n ~ 500,000), we evaluated 513 traits encompassing ICD-10 defined diagnoses and laboratory-derived traits. Our analysis revealed hundreds of significant (FDR < 0.05) associations between genetically regulated cell death gene expression and an array of human phenotypes encompassing both clinical diagnoses and hematologic parameters, which were independently validated in another large-scale DNA biobank (BioVU) at Vanderbilt University Medical Center (n = 94,474) with matching phenotypes. Cell death genes were highly enriched for significant associations with blood traits versus non-cell-death genes, with apoptosis-associated genes enriched for leukocyte and platelet traits. Our findings are also concordant with independently published studies (e.g. associations between BCL2L11/BIM expression and platelet & lymphocyte counts). Overall, these results suggest that cell death genes play distinct roles in their contribution to human phenotypes, and that cell death genes influence a diverse array of human traits.

21 The Microbial Antibiogram as a Function of Testing Indication: Susceptibility Analysis of Escherichia coli from Symptomatic and Asymptomatic Bacteriuria Patients, 2020-2021

OBJECTIVES/GOALS: Antibiograms are used to guide empiric antibiotic selection. However, it is unclear if antibiotic profiles differ between symptomatic urinary tract infections (UTIs) and asymptomatic bacteriuria (ASB). We aimed to compare antibiotic susceptibility profiles of urinary E. coli isolates from patients with a symptomatic UTI to those with ASB. METHODS/STUDY POPULATION: We conducted a cohort study of 1,140 urinary E. coli isolates from unique patients that received care through Vanderbilt University Medical Center (VUMC) from Nov 2020 – Jun 2021. We included any patient that was seen at VUMC as an inpatient, outpatient or at the emergency department with ≥ 105 colony forming units/mL E. coli detected from a clinical urine specimen. Chart abstractions were performed to capture reported UTI symptoms and demographic information. Descriptive statistics were conducted to compare antibiotic susceptibility profiles (i.e., susceptible, intermediate, resistant) between symptomatic and ASB groups. The risk of detection of a multidrug-resistant organism (MDRO) (intermediate, or resistant to at least one antibiotic in three or more classes) was assessed between groups. RESULTS/ANTICIPATED RESULTS: Among 1,140, 1,018 (89%) and 122 (11%) were symptomatic and ASB, respectively. When comparing symptomatic and ASB, the median ages were 50 and 46. Groups had similar proportions of no indwelling catheter (94% v. 95%) and without diabetes (87% v. 88%). The collection setting between inpatient, emergency department, and outpatient were similar with most being outpatient (79% v. 83%). The proportion of patients who were pregnant, immuno compromised, or had a structural/functional urinary tract abnormality were higher in the symptomatic group. The proportion of isolates resistant and susceptible to tested antibiotics were similar between groups, with only ciprofloxacin showing slightly higher resistance among ASB (16% v. 25%). The risk of MDRO detection was similar between groups (RR: 0.858, 95% CI: 0.64, 1.15). DISCUSSION/SIGNIFICANCE: Antibiotic susceptibility comparison demonstrated similar profiles, which suggests antibiogram use as appropriate to guide ASB treatment. Results offer insight on whether traditional methods for assessing antibiotic susceptibility on population-levels could benefit from further refinement by patient-specific clinical parameters.

Discovering clinical drug-drug interactions with known pharmacokinetics mechanisms using spontaneous reporting systems and electronic health records

ObjectiveAlthough the mechanisms behind pharmacokinetic (PK) drug-drug interactions (DDIs) are well-documented, bridging the gap between this knowledge and clinical evidence of DDIs, especially for serious adverse drug reactions (SADRs), remains challenging. While leveraging the FDA Adverse Event Reporting System (FAERS) database along with disproportionality analysis tends to detect a vast number of DDI signals, this abundance complicates further investigation, such as validation through clinical trials. Our study proposed a framework to efficiently prioritize these signals and assessed their reliability using multi-source Electronic Health Records (EHR) to identify top candidates for further investigation. MethodsWe analyzed FAERS data spanning from January 2004 to March 2023, employing four established disproportionality methods: Proportional Reporting Ratio (PRR), Reporting Odds Ratio (ROR), Multi-item Gamma Poisson Shrinker (MGPS), and Bayesian Confidence Propagating Neural Network (BCPNN). Building upon these models, we developed four ranking models to prioritize DDI-SADR signals and cross-referenced signals with DrugBank. To validate the top-ranked signals, we employed longitudinal EHRs from Vanderbilt University Medical Center and the All of Us research program. The performance of each model was assessed by counting how many of the top-ranked signals were confirmed by EHRs and calculating the average ranking of these confirmed signals. ResultsOut of 189 DDI-SADR signals identified by all four disproportionality methods, only two were documented in the DrugBank database. By prioritizing the top 20 signals as determined by each of the four disproportionality methods and our four ranking models, 58 unique DDI-SADR signals were selected for EHR validations. Of these, five signals were confirmed. The ranking model, which integrated the MGPS and BCPNN, demonstrated superior performance by assigning the highest priority to those five EHR-confirmed signals. ConclusionThe fusion of disproportionality analysis with ranking models, validated through multi-source EHRs, presents a groundbreaking approach to pharmacovigilance. Our study's confirmation of five significant DDI-SADRs, previously unrecorded in the DrugBank database, highlights the essential role of advanced data analysis techniques in identifying ADRs.

Back Cover Image

ON THE BACK COVER: Control MDA‐MB‐231 cells seen in confocal (60X objective) max intensity projections with alpha‐tubulin (red), DNA (blue), and Golgi marker GM130 (green) labeled.Credit: Laura A. Zahn (Vanderbilt University Medical Center, Nashville, TN, USA) image

279 Highlighting the Expansion and Reach of the Meharry-Vanderbilt Community Engaged Research Core (CERC) Researcher Training Series

OBJECTIVES/GOALS: * Describe the Community Engaged Research Core (CERC) Researcher Training Program modules and objectives. * Examine the expansion and reach of the CERC Researcher Training Program across various institutions/organizations. METHODS/STUDY POPULATION: Through joint efforts from Meharry Medical College and Vanderbilt University Medical Center, the Community Engaged Research Core (CERC) Researcher Training Program began as a resource to provide CERC post-docs with knowledge of the basic principles of community engaged research (CEnR). It has since expanded to reach researchers at different institutions/organizations. This series examines topics from ethics to dissemination of research findings. The purpose of this project is to describe the CERC Researcher Training Program modules and present the expansion and reach of the training series. We examined both online training requests and series evaluations to determine: (1) participant reach, (2) number of module requests, and (3) purpose for using training modules. RESULTS/ANTICIPATED RESULTS: Since its inception in 2018, the program has expanded to reach 16 institutions/organizations across the United States. On average, 45 researchers register for the training series and approximately 16 researchers participate each year. As time progresses, the number of registrants and attendees continues to increase. To date, there have been 110 online training module requests. The majority (~75%) of participants are requesting content for self-training purposes, while others are using the information to train others in their home institutions/organizations (~25%). Researchers are using modules for a range of reasons including recruitment, career development, and developing their own CEnR training. DISCUSSION/SIGNIFICANCE: CEnR has the ability to make research better in how it is planned, executed and disseminated. Implementing the CERC researcher training series and expanding its reach demonstrates the desire and need to alleviate health disparities through researcher and community partnerships.

The Association between Preoperative Blood Pressures and Postoperative Adverse Events.

The relationship between postoperative adverse events and blood pressures in the preoperative period remains poorly understood. This study tested the hypothesis that day-of-surgery preoperative blood pressures are associated with postoperative adverse events. We conducted a retrospective, observational study of adult patients having elective procedures requiring an inpatient stay between November 2017 and July 2021 at Vanderbilt University Medical Center to examine the independent associations between preoperative systolic and diastolic blood pressures (SBP, DBP) recorded immediately before anesthesia care and number of postoperative adverse events - myocardial injury, stroke, acute kidney injury (AKI), and mortality, while adjusting for potential confounders. We used multivariable ordinal logistic regression to model the relationship. The analysis included 57,389 cases. The overall incidence of myocardial injury, stroke, AKI, and mortality within 30 days of surgery was 3.4% (1,967 events), 0.4% (223), 10.2% (5,871), and 2.1% (1,223), respectively. The independent associations between both SBP and DBP measurements and number of postoperative adverse events were found to be U-shaped, with greater risk both above and below SBP 143 mmHg and DBP 86 mmHg - the troughs of the curves. The associations were strongest at SBP 173 mmHg (adjusted odds ratio [aOR] 1.212 versus 143 mmHg; 95% CI, 1.021 to 1.439; p = 0.028), SBP 93 mmHg (aOR 1.339 versus 143 mmHg; 95% CI, 1.211 to 1.479; p < 0.001), DBP 106 mmHg (aOR 1.294 versus 86 mmHg; 95% CI, 1.003 to 1.17671; p = 0.048), and DBP 46 mmHg (aOR 1.399 versus 86 mmHg; 95% CI, 1.244 to 1.558; p < 0.001). Preoperative blood pressures both below and above a specific threshold were independently associated with a higher number of postoperative adverse events, but the data do not support specific strategies for managing patients with low or high blood pressure on the day of surgery.

Abstract 4781: Evaluation of a TP53 biomarker for lung cancer risk in nasal brushings and peripheral blood

Abstract Background and Purpose: Early detection of lung cancer through annual screening with low-dose computed tomography (LDCT) scans enables potentially curative surgical or stereotactic body radiation therapy (SBRT) treatment and thereby significantly reduces mortality. However, LDCT screening among those currently eligible based primarily on age and smoking criteria yields a large number of false-positive findings and many lung cancers occur among individuals who currently are ineligible. A biomarker that identifies lung cancer risk independent from that caused by age and cigarette smoking might increase specificity and increase the number of people eligible for LDCT screening. Recently we demonstrated that a biomarker measuring TP53 somatic mutations in bronchial brush biopsy specimens from grossly normal airway epithelium was significantly associated with cancer status independent of age and smoking status. Additionally, the TP53 biomarker was synergistic with traditional risk score models (e.g. PLCOM2012 risk score). The goal of this pilot study was to test the TP53 biomarker in nasal brush biopsy specimens, a potential surrogate tissue obtained using less invasive collection methods. Methods: Nasal brushings from 10 lung cancer and 11 control subjects were collected at the University of Toledo and Vanderbilt University Medical Center under approved institutional protocols. Genomic DNA (gDNA) was extracted with the Qiagen AllPrep DNA/RNA kit using a modified protocol. Amplifiable copies of gDNA were measured using quantitative PCR. For each test a sample comprising 25,000 amplifiable copies of gDNA was combined with 50,000 copies of a TP53 internal standard and a complexity control, then PCR amplified, purified and barcoded. Samples were pooled to create two NGS libraries, each with approximately equal numbers of cases and controls, and sequenced on an Illumina MiSeq (Two V3 flow cells, 2 x 250 paired-end). Bioinformatic analysis was conducted using Qiagen CLC Workbench. Variants were called using the Basic Variant Detection Tool. Poisson Exact Test was utilized to determine the significance of endogenous variants relative to internal standard variants at the same base position and a Bonferroni correction was applied to minimize false discovery. Comparison of variant numbers between cases and controls was conducted using an f-test followed by a paired t-test. Results and Conclusions: Consistent with previous results in airway epithelium, in this pilot study the TP53 mutation biomarker measured in nasal brushing was associated with lung cancer (p=0.0085). Measurement of the TP53 biomarker in nasal brush specimens in a larger case-control study currently is underway along with a pilot study in gDNA from peripheral blood. Citation Format: Erin L. Crawford, Daniel J. Craig, Patrick T. Gorman, Mohamed Omballi, Rami Ahmad, Eric L. Grogan, Stephen A. Deppen, James C. Willey. Evaluation of a TP53 biomarker for lung cancer risk in nasal brushings and peripheral blood [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 4781.

Abstract 2244: Accelerated aging associated with cancer characteristics and treatments among breast cancer survivors

Abstract There are currently approximately four million breast cancer (BC) survivors in the US. BC survivors may experience accelerated aging, with greater physical dysfunction and cognitive decline than cancer-free women, potentially due to the detrimental effects of BC and/or its treatments (Tx). However, the aging trajectory after BC diagnosis has not been fully elucidated. A biological age measure, PhenoAge acceleration (PAA), derived from chronological age and nine clinical blood chemistry biomarkers, has been associated with aging-related health outcomes and mortality in the general population. Our study aims to evaluate PAA among BC patients by clinical characteristics, cancer Tx, and medications (chemotherapy and endocrine therapy drugs). The study included 1264 BC patients (age 54.7±11.7) recruited at Vanderbilt University Medical Center in 2004-2021. BC-related information and lab test results were obtained from the tumor registry and electronic health records. The differences in PAA (ΔPAA) by BC stage, grade, subtype, Tx, and medications at multiple time points were evaluated using linear mixed models, with the time interval between diagnosis and PAA assessments as a natural spline, and its interaction with the exposures under study as covariates. Associations of PAA with Tx and medications were estimated with adjustment for other Tx and medications. At 10 years post-diagnosis, stage III/IV (vs 0) and intermediate- and high- (vs low-) grade BC were associated with a higher PAA of 4.48 (p&amp;lt;0.001), 1.26 (p=0.03), and 1.95 (p=0.001), respectively; triple-negative (vs hormone receptor+/Her2-) BC was associated with a lower PAA (ΔPAA=-1.96, p=0.004). Compared with surgery, alone or with radiotherapy, surgery plus chemotherapy was associated with a higher PAA at 1 year (ΔPAA ranging from 1.96, p=0.03 to 4.61, p&amp;lt;0.001), while surgery plus endocrine therapy was associated with a higher PAA at 10 years post-diagnosis (ΔPAA ranging from 1.70, p=0.04 to 2.75, p=0.006). The associations of PAA with most chemotherapy agents diminished over time, reverting to negative or null at 10 years. Antimetabolites, on the other hand, were associated with an increased PAA at 2 years (ΔPAA=2.62, p&amp;lt;0.001) which increased further at 10 years post-diagnosis (ΔPAA=7.26, p&amp;lt;0.001). Endocrine therapy and PAA association at 10 years post-diagnosis was likely driven by aromatase inhibitors (AI) (ΔPAA=1.53, p&amp;lt;0.001). Selective estrogen receptor modulators were inversely associated with PAA at year 10 (ΔPAA=-1.03, p=0.01). Additional adjustment for BC stage, grade, and subtype did not materially change the results for Tx and medications. Our study found that BC patients had accelerated aging up to 10 years post-diagnosis, especially those with stage III/IV and high/intermediate-grade BC. The aging trajectory varied by cancer Tx and medications; antimetabolites and AI might accelerate aging in long term BC survivors. Citation Format: Cong Wang, Jill B. De Vis, Kirsten Nguyen, Brigitte Jia, Mason Alford, Anuradha Bapsi Chakravarthy, Xiao-Ou Shu. Accelerated aging associated with cancer characteristics and treatments among breast cancer survivors [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 2244.

Compliance With NCCN Guidelines for Evaluation and Treatment of Anemia Among Patients With Solid Tumors.

NCCN Guidelines for Hematopoietic Growth Factors recommend evaluation and treatment of anemia in patients with cancer. However, a paucity of data exists regarding compliance with these recommendations. A retrospective cohort study was performed of patients diagnosed with any solid tumor at Vanderbilt University Medical Center from 2008 to 2017. Tumor registry-confirmed cancer cases were identified by ICD-O codes using the Synthetic Derivative database. Anemia was defined as hemoglobin (Hgb) level ≤11 g/dL and graded according to CTCAE version 5.0. Absolute, functional, and possible functional iron deficiency were defined based on NCCN Guidelines. A total of 25,018 patients met inclusion criteria. Median age was 60 years. The most common malignancies were respiratory tract, prostate, and nonprostate urologic (11% each). Among 8,695 patients with Hgb levels available prior to diagnosis, 1,484 (17%) were noted to be anemic proximal to diagnosis. Of the 25,018 patients, 11,019 (44%) were anemic within 6 months of diagnosis. Of these patients, 4,686 (43%) had grade 2 (moderate) anemia and 9,623 (87%) had normocytic anemia. Patients with retroperitoneal/peritoneal cancers had the highest prevalence of anemia (83/110; 75%). A total of 4,125 (37%) underwent any evaluation of their anemia, of whom 1,742 (16%) had iron studies performed and 1,528 (14%) had vitamin B12 or folate studies performed. Fewer than half of patients with anemia received treatment (n=4,318; 39%), including blood transfusion (n=3,528; 32%), oral iron supplementation (n=1,279; 12%), or intravenous iron supplementation (n=97; 1%). Anemia treatment was significantly more frequent as the grade of anemia increased (any treatment among grade 1/mild: 12%; grade 2/moderate: 31%; grade 3/severe: 77%; χ2 [2, n=11,019]=3,020.6; P<.001). Patients with penile and testicular cancers had the highest prevalence of anemia evaluation (n=57; 79%). Anemia is common in patients with solid tumors; yet, compliance with NCCN Guidelines for evaluation and treatment of anemia remains low. There are opportunities to improve compliance with guidelines across the spectrum of cancer care.

Cardiometabolic Improvements After Metabolic Surgery and Related Presurgery Factors.

Metabolic surgery remains the most effective and durable treatment for severe obesity and related metabolic diseases. We examined cardiometabolic improvements after metabolic surgery and associated presurgery demographic and clinical factors in a large multiracial cohort. Included were 7804 patients (20-79 years) undergoing first-time metabolic surgery at Vanderbilt University Medical Center from 1999 to 2022. Pre- and 1-year postsurgery cardiometabolic profiles were extracted from medical records, including body mass index (BMI), blood pressure, blood lipids, glucose, and hemoglobin A1c. The 10-year atherosclerotic cardiovascular disease (ASCVD) risk was estimated per American College of Cardiology/American Heart Association equations. Pre- to postsurgery cardiometabolic profiles were compared by paired t-test, and associated factors were identified by multivariable linear and logistic regression. Among 7804 patients, most were women and White, while 1618 were men and 1271 were Black; median age and BMI were 45 years [interquartile range (IQR): 37-53] and 46.4 kg/m2 (IQR: 42.1-52.4). At 1-year postsurgery, patients showed significant decreases in systolic blood pressure (10.5 [95% confidence interval: 10.1, 10.9] mmHg), total cholesterol (13.5 [10.3, 16.7] mg/dL), glucose (13.6 [12.9, 14.4] mg/dL), hemoglobin A1c (1.13% [1.06, 1.20]), and 10-year ASCVD risk (absolute reduction: 1.58% [1.22, 1.94]; relative reduction: 34.4% [29.4, 39.3]); all P < .0001. Older, male, or Black patients showed less reduction in 10-year ASCVD risk and lower odds of diabetes/hypertension/dyslipidemia remission than younger, female, or White patients. Patients with a history of diabetes, hypertension, dyslipidemia, or cardiovascular disease showed less cardiometabolic improvements than those without. Results were similar with or without further adjusting for weight loss and largely sustained at 2-year postsurgery. Metabolic surgery results in significant cardiometabolic improvements, particularly among younger, female, or White patients and those without comorbidities.

Curating retrospective multimodal and longitudinal data for community cohorts at risk for lung cancer.

Large community cohorts are useful for lung cancer research, allowing for the analysis of risk factors and development of predictive models. A robust methodology for (1) identifying lung cancer and pulmonary nodules diagnoses as well as (2) associating multimodal longitudinal data with these events from electronic health record (EHRs) is needed to optimally curate cohorts at scale. In this study, we leveraged (1) SNOMED concepts to develop ICD-based decision rules for building a cohort that captured lung cancer and pulmonary nodules and (2) clinical knowledge to define time windows for collecting longitudinal imaging and clinical concepts. We curated three cohorts with clinical data and repeated imaging for subjects with pulmonary nodules from our Vanderbilt University Medical Center. Our approach achieved an estimated sensitivity 0.930 (95% CI: [0.879, 0.969]), specificity of 0.996 (95% CI: [0.989, 1.00]), positive predictive value of 0.979 (95% CI: [0.959, 1.000]), and negative predictive value of 0.987 (95% CI: [0.976, 0.994]) for distinguishing lung cancer from subjects with SPNs. This work represents a general strategy for high-throughput curation of multi-modal longitudinal cohorts at risk for lung cancer from routinely collected EHRs.

Why do users override alerts? Utilizing large language model to summarize comments and optimize clinical decision support.

To evaluate the capability of using generative artificial intelligence (AI) in summarizing alert comments and to determine if the AI-generated summary could be used to improve clinical decision support (CDS) alerts. We extracted user comments to alerts generated from September 1, 2022 to September 1, 2023 at Vanderbilt University Medical Center. For a subset of 8 alerts, comment summaries were generated independently by 2 physicians and then separately by GPT-4. We surveyed 5 CDS experts to rate the human-generated and AI-generated summaries on a scale from 1 (strongly disagree) to 5 (strongly agree) for the 4 metrics: clarity, completeness, accuracy, and usefulness. Five CDS experts participated in the survey. A total of 16 human-generated summaries and 8 AI-generated summaries were assessed. Among the top 8 rated summaries, five were generated by GPT-4. AI-generated summaries demonstrated high levels of clarity, accuracy, and usefulness, similar to the human-generated summaries. Moreover, AI-generated summaries exhibited significantly higher completeness and usefulness compared to the human-generated summaries (AI: 3.4 ± 1.2, human: 2.7 ± 1.2, P = .001). End-user comments provide clinicians' immediate feedback to CDS alerts and can serve as a direct and valuable data resource for improving CDS delivery. Traditionally, these comments may not be considered in the CDS review process due to their unstructured nature, large volume, and the presence of redundant or irrelevant content. Our study demonstrates that GPT-4 is capable of distilling these comments into summaries characterized by high clarity, accuracy, and completeness. AI-generated summaries are equivalent and potentially better than human-generated summaries. These AI-generated summaries could provide CDS experts with a novel means of reviewing user comments to rapidly optimize CDS alerts both online and offline.

Natural language processing to identify lupus nephritis phenotype in electronic health records

BackgroundSystemic lupus erythematosus (SLE) is a rare autoimmune disorder characterized by an unpredictable course of flares and remission with diverse manifestations. Lupus nephritis, one of the major disease manifestations of SLE for organ damage and mortality, is a key component of lupus classification criteria. Accurately identifying lupus nephritis in electronic health records (EHRs) would therefore benefit large cohort observational studies and clinical trials where characterization of the patient population is critical for recruitment, study design, and analysis. Lupus nephritis can be recognized through procedure codes and structured data, such as laboratory tests. However, other critical information documenting lupus nephritis, such as histologic reports from kidney biopsies and prior medical history narratives, require sophisticated text processing to mine information from pathology reports and clinical notes. In this study, we developed algorithms to identify lupus nephritis with and without natural language processing (NLP) using EHR data from the Northwestern Medicine Enterprise Data Warehouse (NMEDW).MethodsWe developed five algorithms: a rule-based algorithm using only structured data (baseline algorithm) and four algorithms using different NLP models. The first NLP model applied simple regular expression for keywords search combined with structured data. The other three NLP models were based on regularized logistic regression and used different sets of features including positive mention of concept unique identifiers (CUIs), number of appearances of CUIs, and a mixture of three components (i.e. a curated list of CUIs, regular expression concepts, structured data) respectively. The baseline algorithm and the best performing NLP algorithm were externally validated on a dataset from Vanderbilt University Medical Center (VUMC).ResultsOur best performing NLP model incorporated features from both structured data, regular expression concepts, and mapped concept unique identifiers (CUIs) and showed improved F measure in both the NMEDW (0.41 vs 0.79) and VUMC (0.52 vs 0.93) datasets compared to the baseline lupus nephritis algorithm.ConclusionOur NLP MetaMap mixed model improved the F-measure greatly compared to the structured data only algorithm in both internal and external validation datasets. The NLP algorithms can serve as powerful tools to accurately identify lupus nephritis phenotype in EHR for clinical research and better targeted therapies.

Leveraging generative AI to prioritize drug repurposing candidates for Alzheimer’s disease with real-world clinical validation

Drug repurposing represents an attractive alternative to the costly and time-consuming process of new drug development, particularly for serious, widespread conditions with limited effective treatments, such as Alzheimer’s disease (AD). Emerging generative artificial intelligence (GAI) technologies like ChatGPT offer the promise of expediting the review and summary of scientific knowledge. To examine the feasibility of using GAI for identifying drug repurposing candidates, we iteratively tasked ChatGPT with proposing the twenty most promising drugs for repurposing in AD, and tested the top ten for risk of incident AD in exposed and unexposed individuals over age 65 in two large clinical datasets: (1) Vanderbilt University Medical Center and (2) the All of Us Research Program. Among the candidates suggested by ChatGPT, metformin, simvastatin, and losartan were associated with lower AD risk in meta-analysis. These findings suggest GAI technologies can assimilate scientific insights from an extensive Internet-based search space, helping to prioritize drug repurposing candidates and facilitate the treatment of diseases.