Smoking modulates different secretory subpopulations expressing SARS-CoV-2 entry genes in the nasal and bronchial airways.

SARS-CoV-2 infection and disease severity are influenced by viral entry (VE) gene expression patterns in the airway epithelium. The similarities and differences of VE gene expression (ACE2, TMPRSS2, and CTSL) across nasal and bronchial compartments have not been fully characterized using matched samples from large cohorts. Gene expression data from 793 nasal and 1673 bronchial brushes obtained from individuals participating in lung cancer screening or diagnostic workup revealed that smoking status (current versus former) was the only clinical factor significantly and reproducibly associated with VE gene expression. The expression of ACE2 and TMPRSS2 was higher in smokers in the bronchus but not in the nose. scRNA-seq of nasal brushings indicated that ACE2 co-expressed genes were highly expressed in club and C15orf48+ secretory cells while TMPRSS2 co-expressed genes were highly expressed in keratinizing epithelial cells. In contrast, these ACE2 and TMPRSS2 modules were highly expressed in goblet cells in scRNA-seq from bronchial brushings. Cell-type deconvolution of the gene expression data confirmed that smoking increased the abundance of several secretory cell populations in the bronchus, but only goblet cells in the nose. The association of ACE2 and TMPRSS2 with smoking in the bronchus is due to their high expression in goblet cells which increase in abundance in current smoker airways. In contrast, in the nose, these genes are not predominantly expressed in cell populations modulated by smoking. In individuals with elevated lung cancer risk, smoking-induced VE gene expression changes in the nose likely have minimal impact on SARS-CoV-2 infection, but in the bronchus, smoking may lead to higher viral loads and more severe disease.

Transfer Learning for Automated COVID-19 B-Line Classification in Lung Ultrasound.

Lung ultrasound (LUS) as a diagnostic tool is gaining support for its role in the diagnosis and management of COVID-19 and a number of other lung pathologies. B-lines are a predominant feature in COVID-19, however LUS requires a skilled clinician to interpret findings. To facilitate the interpretation, our main objective was to develop automated methods to classify B-lines as pathologic vs. normal. We developed transfer learning models based on ResNet networks to classify B-lines as pathologic (at least 3 B-lines per lung field) vs. normal using COVID-19 LUS data. Assessment of B-line severity on a 0-4 multi-class scale was also explored. For binary B-line classification, at the frame-level, all ResNet models pretrained with ImageNet yielded higher performance than the baseline nonpretrained ResNet-18. Pretrained ResNet-18 has the best Equal Error Rate (EER) of 9.1% vs the baseline of 11.9%. At the clip-level, all pretrained network models resulted in better Cohen's kappa agreement (linear-weighted) and clip score accuracy, with the pretrained ResNet-18 having the best Cohen's kappa of 0.815 [95% CI: 0.804-0.826], and ResNet-101 the best clip scoring accuracy of 93.6%. Similar results were shown for multi-class scoring, where pretrained network models outperformed the baseline model. A class activation map is also presented to guide clinicians in interpreting LUS findings. Future work aims to further improve the multi-class assessment for severity of B-lines with a more diverse LUS dataset.