2021 Health Equity and Inclusiveness Junior Investigator Award Neel Chudgar, MD Montefiore Medical Center Bronx NY Research Summary Lung cancer remains the most common cause of cancer death in the United States. Underrepresented minorities are at greater risk; Black patients have higher rates of advanced disease and worse survival for all stages. Lung cancer screening has been recommended since 2013, but remains underutilized, especially in minority communities. Care for many at-risk patients is often administered in the emergency department setting, where follow-up is fractured and incomplete. This may be particularly true for incidentally discovered lung nodules, commonly found on imaging done for other purposes, and often diagnosed later as cancers. However, when detected incidentally they are at risk of being lost to follow up with the potential to progress to more advanced stages with a greater risk of mortality. To address this need, we propose to develop an incidental pulmonary nodule management program targeted towards the underserved community cared for by our institution. Through the incorporation of software to automatically detect notation of lung nodules in radiology reports, patients with lung nodules will be captured and integrated into a nodule follow-up clinic. Individual risk for malignancy will be evaluated using artificial intelligence-based calculators derived from the electronic medical record to facilitate guideline concordant care. Finally we will evaluate a blood-based cancer biomarker to further risk stratify patients and will prospectively apply it to patients enrolled in our nodule clinic. Our overall objective is to facilitate the earlier diagnosis of lung cancer in our underserved population, in order to decrease lung cancer mortality. Technical Abstract A common source of lung cancer morbidity is presented by incidental pulmonary nodules (IPNs), which can represent missed cancers when incompletely followed. Underrepresented patients are particularly at risk of guidelines discordant care. Our institution cares for a large proportion of these minority communities. We therefore propose to develop a comprehensive program for the identification, risk stratification and management of IPNs identified through imaging in the emergency setting. We first seek to incorporate lung nodule detection and management software into our electronic medical record (EMR). To facilitate recognition of nodules, we will utilize natural language processing to screen and flag imaging. A multidisciplinary ensemble of practitioners will validate findings and participate in management. At the crux of this team will be a patient navigator to communicate with patients and assist in follow-up. We next seek to create an IPN risk prediction algorithm. Clinical and radiographic elements from the EMR will be subjected to artificial intelligence-based review to automatically leverage data towards a machine learning-derived algorithm. The final element of our proposal will investigate a blood-based gene expression assay as a marker for malignancy. Following its evaluation using a cohort of biobanked samples, we will apply this RNA-based detection assay to look for markers of malignancy to facilitate the early detection of lung cancers. We ultimately seek to improve lung cancer care within our largely underserved population through a combination of incorporating patients with IPNs into the health care system and providing appropriate and automated risk stratification to expedite care and diagnosis.
