Daily Respiratory Research Analysis
Three impactful respiratory studies stood out: an interpretable dual-phase CT AI (NeoPred) that predicts pathological response to neoadjuvant chemo-immunotherapy in resectable NSCLC; a multi-trait genetic analysis identifying 31 new asthma loci in East Asians and implicating a CD36 missense variant; and a large cohort linking pulmonary arterial pressure trajectories before/after kidney transplant to mortality, spotlighting post-transplant pulmonary hypertension risk.
Summary
Three impactful respiratory studies stood out: an interpretable dual-phase CT AI (NeoPred) that predicts pathological response to neoadjuvant chemo-immunotherapy in resectable NSCLC; a multi-trait genetic analysis identifying 31 new asthma loci in East Asians and implicating a CD36 missense variant; and a large cohort linking pulmonary arterial pressure trajectories before/after kidney transplant to mortality, spotlighting post-transplant pulmonary hypertension risk.
Research Themes
- AI-driven imaging biomarkers for thoracic oncology
- Population-specific asthma genetics and immune mechanisms
- Hemodynamic trajectories of pulmonary hypertension and transplant outcomes
Selected Articles
1. NeoPred: dual-phase CT AI forecasts pathologic response to neoadjuvant chemo-immunotherapy in NSCLC.
In multi-center cohorts (retrospective n=509; prospective n=50), a dual-phase CT 3D-CNN (NeoPred) predicted major/pathologic complete response to neoadjuvant chemo-immunotherapy in resectable NSCLC, outperforming unaided radiologists and improving their performance when used as assistance. External validation AUC reached 0.772 (0.787 with clinical variables) and prospective AUC 0.760; expert AUC improved to 0.829 with model assistance.
Impact: Provides an interpretable, non-invasive biomarker to guide surgical timing and escalation/de-escalation after neoadjuvant chemo-immunotherapy in NSCLC, surpassing expert assessment and generalizing across centers.
Clinical Implications: May enable personalized surgical decision-making, triage for additional cycles vs early surgery, and stratification for trials; facilitates resource allocation in radiology and pathology by anticipating response.
Key Findings
- Dual-phase CT 3D-CNN (NeoPred) achieved AUC 0.772 in external validation, rising to 0.787 with clinical variables.
- Prospective cohort performance was AUC 0.760, exceeding the mean expert AUC of 0.720.
- Model assistance improved pooled expert AUC to 0.829 and accuracy to 0.820.
- Robust performance persisted in radiologic stable-disease subgroups (AUC up to 0.833).
Methodological Strengths
- Multi-center cohorts with external and prospective validation
- Head-to-head comparison with nine board-certified radiologists and interpretable analysis (SHAP)
Limitations
- Retrospective development on Chinese cohorts may limit global generalizability
- No randomized clinical utility assessment or impact on outcomes
Future Directions: Prospective, multinational impact trials to assess decision-change and outcomes; integration with PET/biopsy and circulating biomarkers to form multi-modal predictors.
2. Multi-trait genetic analysis of asthma and eosinophils uncovers pleiotropic loci in East Asians.
In East Asians, multi-trait GWAS linked asthma to eosinophil counts (LDSC and MR) and identified 52 loci, including 31 previously unreported. A missense variant in CD36 (rs75326924) showed higher expression in lymphocytes and ILC2-enriched cells in asthma and associated proteomic changes (downregulated IL-7, Oncostatin M, VEGFA), suggesting CD36-related immune pathways as population-specific targets.
Impact: Expands asthma genetics in an underrepresented population and pinpoints a functionally supported CD36 variant with immunologic readouts, informing precision therapeutics for East Asians.
Clinical Implications: Supports development of population-tailored biomarkers and therapies (e.g., targeting CD36/ILC2 pathways) and refines risk prediction models for East Asians.
Key Findings
- LDSC and MR demonstrated significant genetic correlation between asthma and eosinophil count in East Asians.
- MTAG identified 52 genome-wide significant loci, including 31 novel loci specific to East Asians.
- A missense variant in CD36 (rs75326924) showed increased expression in lymphocytes and ILC2-enriched cells and was associated with downregulation of IL-7, Oncostatin M, and VEGFA.
- Findings highlight pleiotropy across asthma and WBC traits and nominate CD36-related pathways as therapeutic targets.
Methodological Strengths
- Integration of LDSC, Mendelian randomization, and MTAG across East Asian cohorts
- Functional validation via flow cytometry and proteomics
Limitations
- Population focus on East Asians may limit generalizability to other ancestries
- Observational genetic associations without interventional validation
Future Directions: Replication in multi-ancestry cohorts, mechanistic dissection of CD36-ILC2 signaling, and early-phase trials targeting implicated pathways.
3. Longitudinal Pulmonary Arterial Pressure Trajectories and Clinical Outcome in Kidney Transplantation Patients.
In 631 kidney transplant recipients with pre- and post-transplant echocardiographic PAPs, new-onset or persistent pulmonary hypertension after transplant independently increased mortality versus no PH, whereas resolved PH showed the most favorable survival. Each 10 mmHg increase in sPAP from pre- to post-KT raised mortality risk by 21%, and a 10 mmHg decrease reduced risk by 17%; findings were validated in an external cohort.
Impact: Establishes PAP trajectory as a dynamic biomarker for post-transplant risk stratification, highlighting modifiable hemodynamic targets that correlate with survival.
Clinical Implications: Supports integrating serial PAP assessment into KT evaluation and follow-up to identify high-risk patients (New/Persistent-PH) and prioritize PH management to improve survival.
Key Findings
- New-PH and Persistent-PH post-transplant increased mortality risk vs No-PH (HR 1.51 and 1.37, respectively).
- Resolved-PH had the most favorable survival; trend toward lower risk vs No-PH.
- Each 10 mmHg increase in sPAP from pre- to post-KT increased mortality by 21%; a 10 mmHg decrease reduced mortality by 17%.
- Results reproduced in a sex-balanced validation cohort using echo and right heart catheterization.
Methodological Strengths
- Large, real-world cohort with pre- and post-transplant measurements
- Independent validation cohort and consistent findings across modalities
Limitations
- Retrospective design with potential residual confounding
- Echocardiographic sPAP as a surrogate may misclassify PH severity
Future Directions: Prospective interventional studies to test PH optimization pre/post-KT and determine thresholds/targets for sPAP that improve survival.