Clinical Scorecard: Creation and assessment of a CT-radiomics framework for diagnosing lung cancer linked to cystic airspaces: a multicenter investigation

At a Glance

Key Highlights

• LCCA is a rare lung malignancy often misdiagnosed due to cystic airspaces previously considered benign.
• Needle biopsy is challenging and risky in LCCA due to cystic structures; non-invasive diagnostic methods are needed.
• Radiomics using CT imaging and clinical features can accurately differentiate malignant from benign cystic lung lesions.

Guideline-Based Recommendations

Diagnosis

• Use preoperative chest CT to identify lung lesions with cystic airspaces >5 mm diameter.
• Apply radiomics feature extraction from tumor and peritumor regions on CT images for diagnostic assessment.
• Involve multiple radiologists for independent CT image assessment and resolve discrepancies by expert review.

Management

• Consider surgical resection for lesions classified as malignant by radiomics and clinical evaluation.
• Avoid reliance on needle biopsy alone due to increased pneumothorax risk in cystic lesions.

Monitoring & Follow-up

• Monitor radiologic features such as lesion diameter, location, ground-glass opacity, spicule sign, pleural indentation, and emphysema presence.
• Use radiomics model outputs to guide follow-up and treatment decisions.

Risks

• Risk of pneumothorax with needle biopsy in cystic airspace lesions.
• Potential misdiagnosis if cystic airspaces are assumed benign without advanced imaging analysis.

Patient & Prescribing Data

Patients undergoing evaluation for lung lesions with cystic airspaces on CT imaging

Radiomics-based non-invasive diagnosis can improve accuracy and reduce biopsy-related complications, guiding appropriate surgical intervention.

Clinical Best Practices

• Define cystic airspaces as air-containing structures with walls and diameter >5 mm on CT.
• Use a 5-mm peritumoral margin for radiomics feature extraction to capture relevant tumor environment characteristics.
• Standardize CT image acquisition and preprocessing including voxel resampling and intensity normalization.
• Employ multi-reader assessment and consensus for radiologic feature evaluation to enhance diagnostic reliability.

References

• Fleischner Society definition of lung cystic airspace
• Observational study on missed carcinomas presenting with cystic airspaces
• Radiomics predictive capabilities in NSCLC diagnosis and prognosis
• FAE software for radiomics feature extraction