PREVALENCE OF INTERSTITIAL LUNG ABNORMALITIES AND ASSOCIATION WITH MORTALITY AND HOSPITALIZATION IN A CT LUNG CANCER SCREENING COHORT

Abstract

TOPIC: Diffuse Lung Disease TYPE: Original Investigations PURPOSE: Interstitial lung abnormalities (ILA), non-dependent interstitial changes identified on lung CT scans, have been shown to have significant biological and clinical consequences. In some cases, ILA may represent a less severe or earlier stage of pulmonary fibrosis. In the United States, 14.5 million patients are estimated to be eligible for CT lung cancer screening (CTLS) which has the potential to be an early screening tool for interstitial lung disease. In this study, we sought to characterize ILA in a large CTLS cohort, evaluating prevalence and associations with clinically relevant outcomes. METHODS: Patients (n=1703) from Lahey Hospital and Medical Center who underwent CTLS from Jan 12, 2012 through Sep 30, 2014 and had an in network primary care physician were included; outcomes data was collected through Sep 30, 2019. Two thoracic radiologists certified in lung cancer screening scored the baseline CT exams per Fleischner Society ILA position paper guidelines (ILA, indeterminate, no ILA). For CT scans in which the score was discordant between the two radiologists (n=163), a third reader scored the CT to determine the final score. Cox multivariable regression proportional hazards models were used to test the association between ILA and cancer, all-cause mortality, all cause hospitalization and COPD and pneumonia related hospital admission adjusted for age, sex, smoking status and pack years exposure. Kaplan-Meier plots were generated to visualize the associations between observed baseline ILA and clinical outcomes. To adjust for multiplicity Bonferroni correction was utilized and p-value significance was set at<=0.01. All statistical analyses were performed using STATA14.1. RESULTS: Baseline characteristics were mean age 62.6 ± 6.2 years, 56.3% male, 57.7% with emphysema. A total of 101 (5.9%) patients were scored indeterminate and 45 (2.6%) were scored for ILA on the baseline CT. Of those with ILA, 35 (78%) had not been seen by a pulmonologist prior to their baseline CT and average time between baseline CTLS exam and referral was 2.51 ± 2.85 years. Of the patients with baseline ILA, who did not carry a pre-scan diagnosis of interstitial lung disease (ILD) (n=39), 15(33.3%) subsequently received a clinical diagnosis of ILD, 5.0 ± 2.5 years post baseline scan. In multivariable modeling, the presence of ILA was associated with death, HR 4.05 (2.26, 7.23, p<0.001) and hospital admission for pneumonia, HR 3.07 (1.41, 6.68, p<0.005). CONCLUSIONS: The results of this study demonstrate that ILA are a significant clinical finding in a large CTLS cohort with a prevalence and 5-year mortality risk similar to that of lung cancer. A significant percentage of these patients develop clinical ILD and may benefit from early pulmonology referral, evaluation, and management. CLINICAL IMPLICATIONS: Identification of ILA in a CTLS cohort has the potential to identify patients who will develop clinical ILD. DISCLOSURES: No relevant relationships by Melissa Gawlik, source=Web Response Grant funding relationship with Genentech Please note: 9/10/2020 Added 04/23/2021 by Lee Gazourian, source=Web Response, value=Grant/Research Support No relevant relationships by Jeffrey Hashim, source=Web Response No relevant relationships by Gary Hunninghake, source=Web Response Advisor relationship with AstraZeneca Please note: 2019-2021 Added 04/22/2021 by Brady McKee, source=Web Response, value=Advisor Fee No relevant relationships by Timothy MD, source=Web Response No relevant relationships by Ezra Miller, source=Web Response Grant recipient relationship with Genentech Please note: 07/01/2020 -- Added 04/18/2021 by Avignat Patel, source=Web Response, value=Grant/Research Support No relevant relationships by Lori Lyn Price, source=Web Response No relevant relationships by Shawn Regis, source=Web Response No relevant relationships by Christoph Wald, source=Web Response, value=Honoraria Removed 04/22/2021 by Christoph Wald, source=Web Response