Presenting symptoms, screening and diagnosis in asian and white patients with stage IV lung cancer.

Abstract

e18687 Background: Lung cancer (LC) is a leading cancer diagnosis and cause of cancer death in Asian Americans, the fastest growing racial/ethnic group in the US. Some racial/ethnic groups are less likely to be eligible for guideline-recommended LC screening with low-dose computed tomography (LDCT). Data on presenting symptoms (PS) and initial evaluation in Asian patients (AP) with stage IV LC in the US are limited. This retrospective study examines LDCT utilization and patterns of PS for AP and White patients (WP) with stage IV LC at an urban academic medical center serving a large proportion of AP. Methods: Patients newly diagnosed with stage IV LC from 01/01/2014 to 12/31/2019 were identified from Tufts Medical Center cancer registry. Demographics, clinical characteristics, and biomarker data were compared between AP and WP by Mann–Whitney U test and Chi-square/Fisher's exact tests for continuous and categorical variables, respectively. Univariable and multivariable logistic regressions were used to assess associations between PS and race. Results: Of 89 AP and 197 WP, AP had significantly more non-small cell (NSC) histology (85 vs 73%), never-smokers (36 vs 7%) and male representation (72 vs 48%). Similar rates of brain metastasis (35 vs 39%) were observed. Among 219 patients with NSCLC, AP had more driver mutations compared to WP (53 vs 29%), particularly more EGFR alterations (68 vs 15% of observed driver mutations). Similar PD-L1 positivity (34 vs 32%) was noted at diagnosis in AP and WP with stage IV NSCLC. At diagnosis, fewer AP (N = 14) vs WP (N = 47) met USPSTF 2013 eligibility criteria for LDCT screening (16 vs 24%, p = 0.162), however 21% (N = 3) of eligible AP completed LDCT screening compared to 9% (N = 4) of eligible WP (p = 0.336). AP tended to have more primary care (PC) visits prior to diagnosis (51 vs 38%, p = 0.064), but longer median duration of symptoms prior to diagnosis (8 vs 4 weeks, p = 0.003). More AP presented with respiratory symptoms (80 vs 54%, p < 0.001), namely cough (58 vs 33%, p < 0.001), chest pain (20 vs 8%, p = 0.004) and hemoptysis (15 vs 7%, p = 0.050). Fewer AP presented with symptoms due to metastatic disease (34 vs 51%, p = 0.011). Rates of constitutional symptoms (42 vs 35%, p = 0.311) and asymptomatic diagnoses (3.4 vs 3.0%) were similar. Adjusting for age, sub-stage, histology, and smoking status, multivariable logistic regression showed significant association between AP and respiratory symptoms presentation (OR = 3.43, p < 0.001). Conclusions: USPSTF 2013 LC screening criteria did not capture most of the AP and WP diagnosed with stage IV LC in this study and missed racial nuances in LC screening. Although limited in size, this study found that AP, more of whom are nonsmoking, are diagnosed with stage IV LC after longer preceding often respiratory symptoms, despite similar PC utilization and LDCT uptake as compared to WP. Awareness of race and associated PS should inform assessment for LC screening in AP.