A Multicenter Study in Bronchoscopy Combining Stimulated Raman Histology With Artificial Intelligence for Rapid Lung Cancer Detection - The ON-SITE Study

Condition(s) studied

Study summary

The ON-SITE study represents a prospective, observational study focused on the training/tuning and pivotal validation of deep learning algorithms that detect cell/tissue morphology suspicious for cancer in biopsies of peripheral lung nodules/masses and mediastinal/hilar lymph nodes imaged with the NIO Laser Imaging System in the procedure room without requiring traditional sample processing.

The study includes four arms based on biopsy location and biopsy modality/tool:

1. Transbronchial forceps biopsy of peripheral lung nodules/masses (peripheral-TBBx)
2. Transbronchial needle aspiration biopsy of peripheral lung nodules/masses (peripheral TBNA)
3. Transbronchial needle aspiration biopsy of mediastinal/hilar lymph nodes (EBUS-TBNA)
4. Transbronchial cryo biopsy of peripheral lung nodules/masses (peripheral-CBx)

Eligibility

Primary outcome measure(s)

• Peripheral TBBx and CBx Co-primary Endpoints — From enrollment to the End of the Procedure.
  Detection of lung cancer in peripheral TBBx and CBx biopsies by NIO image analysis module with non-inferior performance compared to rapid on-site evaluation touch imprint cytology (ROSE-TIC)
• Peripheral Lung Transbronchial Fine Needle Aspiration Endpoint — From enrollment to the End of the Procedure.
  Detection of lung cancer in peripheral lung transbronchial needle aspiration by NIO image analysis module with non-inferior performance compared to rapid on-site evaluation (ROSE).
• Lymph Node Endobronchial Ultrasound Transbronchial Fine Needle Aspiration Endpoint — From enrollment to the End of the Procedure.
  Determine area-under-the-curve (AUC) of the receiver-operating-characteristic (ROC) of NIO image analysis module for detection of cancer at a lymph-node location.

Trial sites (7)