Lung cancer and pulmonary infections including bacterial, fungal, and tuberculosis
Key Mechanisms
Multimodal machine learning analysis of microbial profiles, host gene expression, immune cell composition, and tumor fraction from bronchoalveolar lavage fluid metagenomic next-generation sequencing (mNGS) data
Target Population
Patients with lung lesions suspected of lung cancer or pulmonary infections
Care Setting
Clinical diagnostic laboratories and pulmonology care settings
Key Highlights
mNGS enables simultaneous detection of pathogens and host genetic information from bronchoalveolar lavage fluid within 24 hours.
Integrated host/microbe machine learning model (Model VI) differentiates lung cancer from bacterial, fungal, and tuberculosis infections with high accuracy (AUC up to 0.937).
Rule-in/rule-out diagnostic strategy improves accuracy for distinguishing lung cancer from specific pulmonary infections, facilitating timely and cost-effective clinical decision-making.
Guideline-Based Recommendations
Diagnosis
Utilize bronchoalveolar lavage fluid mNGS to obtain comprehensive microbial and host transcriptomic data.
Apply integrated machine learning models combining microbial profiles, host gene expression, immune cell composition, and tumor fraction for differential diagnosis.
Employ rule-in/rule-out strategies to enhance diagnostic accuracy between lung cancer and pulmonary infections.
Management
Use rapid mNGS-based diagnosis to guide appropriate treatment selection for lung cancer versus infectious etiologies.
Minimize reliance on multiple invasive or time-consuming tests by leveraging mNGS data.
Monitoring & Follow-up
Monitor host immune response and microbial community changes via mNGS to assess treatment response and disease progression.
Risks
Potential misdiagnosis if relying solely on clinical and radiological features without molecular diagnostics.
Delayed or inappropriate treatment due to overlapping symptoms and imaging findings between lung cancer and infections.
Patient & Prescribing Data
Patients with lung lesions undergoing evaluation for lung cancer or pulmonary infections
Early and accurate differentiation using mNGS-guided diagnostics supports timely initiation of targeted therapies, reducing morbidity associated with misdiagnosis.
Clinical Best Practices
Collect bronchoalveolar lavage fluid samples for mNGS testing in patients with ambiguous lung lesions.
Integrate microbial and host genomic data using validated machine learning models for differential diagnosis.
Implement rapid turnaround times (within 24 hours) for mNGS results to inform clinical decisions.
Consider patient clinical features alongside mNGS data to improve diagnostic confidence.
Use mNGS as a cost-effective tool to reduce the need for multiple diagnostic procedures.
