C34-31 Nasal Epithelial Transcriptomic Profiles Identify Distinct Asthma Endotypes in a Study of Puerto Rican and Dominican Adults
S Yang, Y -Y Han, A Pirzada, R Kaplan, F J Rosser, E Forno, G Hudes, W Chen, C Isasi, J C CeledonAbstract
Rationale
Comprehensive categorization of asthma endotypes has previously required a bronchoscopy. Whether nasal epithelial transcriptomic profiles can identify asthma endotypes in adults, including Latino adults, is unclear.
Methods
We analyzed nasal epithelial transcriptomic profiles from 648 Puerto Rican and Dominican adults with (cases, n=312) and without (controls, n=336) asthma aged 29-90 years who were recruited at the Bronx (NY) and Chicago (IL) sites of the Hispanic Community Health Study/Study of Latinos (HCHS/SOL). At the baseline visit, asthma was defined as having had asthma diagnosed by a physician (physician-diagnosed asthma). Using thirteen “signature genes” for T helper (T)2, T17, and T1 immunity, we classified transcriptomic profiles into five groups, corresponding to five endotypes: T2HIGH, T17HIGH, T1HIGH, T17HIGH/T1HIGH, and T2LOW/T17LOW/T1LOW. K-means and Gaussian mixed modeling were used for robust clustering. Accounting for the complex survey design of HCHS/SOL, multinomial logistic regression was used to compare participants in each asthma endotype against controls (the reference group). All models were adjusted for age, sex, body mass index (BMI), study site (Bronx vs. Chicago), self-reported Hispanic subgroup (Puerto Rican vs. Dominican), birthplace in the U.S. vs. elsewhere, smoking status (never, former, and current), pack-years of cigarette smoking, physician-diagnosed chronic obstructive pulmonary disease (COPD), percent predicted (%pred) FEV1 and %pred FEV1/FVC.
Results
In 312 cases, we identified five asthma endotypes: T2HIGH (n=16 [5.1%]), T17HIGH (n=96 [30.8%]), T1HIGH (n=71 [22.8%]), T17HIGH/T1HIGH (n=53 [17%]), and T2LOW/T17LOW/T1LOW (n=76 [24.3%]) (see Figure). In a multivariable analysis with controls as the reference group, female sex was associated with increased odds of T1HIGH, T17HIGH/T1HIGH, and T2LOW/T17LOW/T1LOW asthma; Puerto Rican ethnicity was associated with T17HIGH asthma (OR = 2.7, 95%CI=1.1-6.2); COPD was associated with T17HIGH asthma (OR = 3.9, 95%CI=1.0-15.5) and T17HIGH/T1HIGH asthma (OR = 8.1, 95%CI=2.1-31.9), current smoking was associated with T2LOW/T17LOW/T1LOW asthma (OR = 5.8, 95%CI=2.3-14.7), and FEV1/FVC was associated with lower odds of each endotype except T2LOW/T17LOW/T1LOW asthma. Similar findings were obtained in a sensitivity analysis adjusting for the top principal components derived from genome-wide genotypic data instead of Hispanic subgroup.
Conclusions
Nasal epithelial transcriptomic profiles identified five distinct asthma endotypes in Puerto Rican and Dominican adults. Consistent with our prior findings in minority children, T2HIGH asthma was less common than all other (“T2-low”) endotypes. Puerto Rican adults, known to have higher asthma morbidity and mortality, were more likely to have T17-high asthma. COPD was associated with T17HIGH and T17HIGH/T1HIGH asthma, and thus asthma-COPD overlap may be more likely in Latino adults with T17HIGH asthma endotypes.
This abstract is funded by: HL152475, HL129949