Bhatt, S

Abstract Rationale In fungal endemic regions, differentiating benign from malignant pulmonary nodules on CT scans remains a major diagnostic challenge. Coccidioides nodules frequently resemble malignant lesions on CT and PET imaging, presenting significant challenges for differentiation, even among radiologists. Consequently, the benign biopsy and resection rate, especially in the Southwestern US, remains high. We previously developed [DOI:10.1371/journal.pone.0196910] and externally validated [DOI:10.1183/13993003.02485-2020] the Benign versus aggRessive nODule Evaluation using Radiomic Stratification (BRODERS) machine learning (ML) classifier to non-invasively distinguish between benign and malignant pulmonary nodules with an AUC of 0.87. However, its performance in certain benign nodules with malignant surface and texture characteristics like in coccidioidomycosis had a suboptimal AUC of 0.65. In this study, we redesigned our radiomics model by including numerous clinical features. Methods In this retrospective study, 113 pulmonary nodules (66 Coccidioides and 47 adenocarcinoma spectrum) detected on CTs with ground truth pathology diagnosis were identified by screening for key terms like “nodule”, “cocci”, and “adenocarcinoma” from both CT and pathology reports using a large language model trained on clinical textual data (Illuminate®). Clinical features such as demographics, smoking history, and laboratory studies were extracted by trained clinicians from the electronic medical record. All nodules were segmented and radiomic features extracted using a semi-automated seed growing algorithm previously described [DOI: 10.1097/JTO.0b013e3182843721]. The database was randomly split into an 80%-20% development and hold-out set prior to ML modeling. Feature selection for variables was conducted utilizing Elastic Net Regression, which provides a balanced approach between Lasso and Ridge regression. A logistic regression model was developed using the selected features and ground truth information of the training cohort. The true performance of the model was tested on the hold-out set. Results 160 radiomics and 22 clinical features were extracted. Elastic Net regression identified 44 features that contributed the most. Some of these clinical features include presence of satellite lesions, smoking history, and presence of positive IgG coccidiomycosis serology. The AUC of the ROC curve for the hold-out test set was 0.86. Conclusions The clinical features of significance that emerged during multimodal modeling are explainable. Designing explainable ML models is important, especially in the Southwestern US where Coccidioides nodules can appear to be malignant to both human experts and ML models. Although coccidioidomycosis serologies alone lack sensitivity and specificity to diagnose Coccidioides nodules, integrating them with a multimodal ML model can improve differentiation between benign Coccidioides and malignant adenocarcinoma nodules. This abstract is funded by: None

Abstract Rationale Patients with idiopathic pulmonary fibrosis (IPF) living in disadvantaged neighborhoods have worse composite outcomes and pulmonary function. Although the current FDA approved drugs reduce lung function decline and may improve mortality for this progressive and fatal disease, equitable access to therapy remains a challenge, which may be contributing to low antifibrotic uptake. Existing understanding of social determinants of health in IPF is derived from local registries and special subpopulations, such as Veterans. The aim of our study is to investigate the association between neighborhood-level deprivation indicators and both the distribution of patients with IPF and outcomes in a large national database. Methods We performed a retrospective cohort analysis between October 1st, 2014, and December 31st 2022, using the Optum Labs Data Warehouse, a large de-identified administrative claims database. We used claims data from commercial, Medicare Advantage, and Medicare Fee-for-Service enrollees across the United States. Diagnosis of IPF was defined using ICD-9 and ICD-10 codes. Social deprivation index (SDI), assessed at the zip code tabulation area level, was determined based on 7 demographic characteristics from the 2015-2019 American Community Survey. Patients were grouped in quintiles based on SDI at diagnosis, where higher values reflect greater deprivation. Hazard ratios (HR) were calculated for all-cause mortality as primary and hospitalization as secondary outcome. Results A total of 96,143 patients (mean age 75.1 years ± 6.9 years) were included in the study, 23.5% received antifibrotics. As area deprivation increased, proportion of female, Black, and Hispanic patients increased, whereas, proportion of male, White, and Asian patients decreased. Similarly, antifibrotic use declined slightly, while prescription of steroids and oxygen increased with SDI. Residing in most disadvantaged neighborhoods increased risk of hospitalization (HR, 1.13 [95% CI 1.09 - 1.17] and all-cause mortality (HR, 1.15 [95% CI 1.11 - 1.19]) after adjusting for covariates over median follow-up of 301 days and 680 days, respectively. Conclusion Gender and racial disparities were evident among IPF patients who live in resource-limited areas. Our study shows that greater proportion of female, Black, and Hispanic IPF patients were disproportionately represented in more socioeconomically deprived areas. Furthermore, as area-level disadvantage increased, fewer patients received antifibrotics, while more patients were on steroids and oxygen therapy potentially indicating more advanced disease. Moreover, living in disadvantaged neighborhoods was associated with worse all-cause mortality and hospitalizations. Identifying individuals who live in socioeconomically disadvantaged areas is essential to provide equitable care and improve outcomes for this life-threatening disease. This abstract is funded by: None

Abstract Background Idiopathic pulmonary fibrosis (IPF) is a chronic progressive fibrosing interstitial lung disease that continues to have high morbidity and mortality. There are only 3 approved medications, (nintedanib, pirfenidone and nerandomilast) which marginally slow FVC decline and only nerandomilast having some mortality benefit. All three have no effect on quality of life (exacerbations and hospitalizations). There is an urgent need to find more effective treatment options. Since IPF is predominantly a disease of the elderly with a peak prevalence in those 65 and older, it is common for these individuals to have other comorbidities such as Hyperlipidemia (HLD). Interestingly, in a prior small study of IPF patients, there is evidence to suggest clinical benefit from ezetimibe, a medication used to treat HLD. This study aims to further evaluate the clinical benefits of ezetimibe in patients with IPF in a large national cohort. Methods We conducted this study using the Optum Labs Data Warehouse (OLDW). The OLDW contains de-identified administrative claims data, including medical and pharmacy claims and enrollment records. This study used claims data from commercial, Medicare Advantage (MA), and Medicare Fee-for-service (FFS) enrollees between October 1st, 2014, and December 31st, 2021. Diagnosis of IPF was defined using ICD-9 and ICD-10 codes. Patients without a diagnosis of IPF were excluded. P < 0.05 was set to determine statistical significance. Two matched patient cohorts depending on ezetimibe use were defined. Primary outcome was all-cause mortality, and secondary outcome was hospitalizations. Results We identified 7724 patients with IPF. After adjusting for age, age group, gender, race, health plan, year of IPF diagnosis, smoking, diabetes, statin treatment, and antifibrotic treatment, two cohorts were identified and matched, patients treated with ezetimibe (n = 3862) and patients not treated with ezetimibe (n = 3862). Among the patients with IPF treated with ezetimibe, ezetimibe use was associated with a reduction in all-cause mortality (hazard ratio [95% CI]: 0.704 [0.660 - 0.752], p-value: <0.001) and hospitalizations(hazard ratio [95% CI]: 0.907 [0.857 - 0.959], p-value: 0.001) compared to patients not treated with ezetimibe. Conclusion In patients with IPF, treatment with ezetimibe may be associated with improved clinical outcomes. Further studies are necessary to confirm benefits, assess anti-fibrotic effects, and inform implementation in management of IPF. As IPF remains a disease with limited treatment options, repurposing drugs continues to offer an avenue to find affordable, safe and already approved medications for use. This abstract is funded by: None