This phase II trial tests the effectiveness and safety of artificial intelligence (AI) to determine dose recommendation during stereotactic body radiation therapy (SBRT) planning in patients with primary lung cancer or tumors that has spread from another primary site to the lung (metastatic). SBRT uses special equipment to position a patient and deliver radiation to tumors with high precision. This method may kill tumor cells with fewer doses over a shorter period and cause less damage to normal tissue. Even with the high precision of SBRT, disease persistence or reappearance (local recurrence) can still occur, which could be attributed to the radiation dose. AI has been used in other areas of healthcare to automate and improve various aspects of medical science. Because the relationship of dose and local recurrence indicates that dose prescriptions matter, decision support systems to help guide dose based on personalized prediction AI algorithms could better assist providers in prescribing the radiation dose of lung stereotactic body radiation therapy treatment.
Patients with radiographic findings consistent with lung cancer or solitary or oligometastatic disease in the lung. Most patients will have primary non-small cell lung cancer. For primary lung cancers, we include lesions with ground glass opacities with a solid component of 50% or greater. Patients with solitary or oligo-metastatic disease in the lung may have any other histology or cancer type.
Patients must have radiographically measurable or evaluable disease per RECIST v. 1.1.
Patients must be age >= 18 years.
Patients must exhibit an Eastern Cooperative Oncology Group (ECOG) performance status of 0-2.
Patients who have had prior radiotherapy with radiation field overlap.
For primary lung cancers, patients with ground glass opacities without a solid component will be excluded.
Patients who have not recovered from adverse events confined to the thorax (i.e. pneumonitis, bronchial insufficiency, bronchial hemorrhage, esophagitis) due to prior anticancer therapy (i.e., have residual toxicities >= grade 2) with the exception of alopecia. low blood counts (neutropenia, anemia, etc), or anatomically distinct toxicities (i.e. cystitis).
Patients who are receiving any other concurrent investigational agents or genotoxic chemotherapy for cancer treatment.
Single group
Computed Tomography
Procedure/SurgeryUndergo CT
Magnetic Resonance Imaging
Procedure/SurgeryUndergo MRI
Positron Emission Tomography
Procedure/SurgeryUndergo PET
Quality-of-Life Assessment
Other interventionAncillary studies
Radiology, Treatment Planning
Other interventionUndergo radiation planning with AI-directed analysis for dose recommendation
Stereotactic Body Radiation Therapy
RadiationUndergo SBRT
X-Ray Imaging
Procedure/SurgeryUndergo x-ray imaging
Local failure free survival (LFS)
LFS is defined as the length of time after SBRT that the patient survives without local failure (as assessed by tumor imaging). LFS data will be collected from the end of SBRT, until the patient experiences recurrence, completes the 5-year follow-up after SBRT, or experiences death from any cause (whichever is sooner). The LFS at 2 years after completion of SBRT will be calculated based on Kaplan-Meier estimates
Progression-free survival (PFS)
For PFS analysis, disease progression is defined as progressive disease per Response Evaluation Criteria in Solid Tumors version 1.1.
Respiratory function
To assess the respiratory toxicity of SBRT doses as recommended by Deep Profiler +iGray, this endpoint will collect and report the frequency of adverse events by type, severity (grade), timing, and attribution, according the RTOG Pulmonary Function Test Toxicity Scale.
Incidence of adverse events
To assess the Toxicity Profile of SBRT fractions as recommended by Deep Profiler +iGray, this endpoint will collect and report the frequency of adverse events by type, severity (grade), timing, and attribution, according the National Cancer Institute-Common Terminology Criteria for Adverse Events version 5.0.
Adherence to recommended dose
To evaluate if individualized radiation doses recommended by machine learning software (Deep Profiler +iGray) can be implemented in a clinical practice (feasibility), we will assess the adherence of the prescribing physician to the dose recommended by the Deep Profiler +iGray during the 1-2 week period of SBRT. The adherence to the dose recommended by the Deep Profiler +iGray will be reported.
Click on the lead sponsor to view all of their trials.
Northwestern University
Lead sponsor
Varian Medical Systems
Collaborator
National Cancer Institute (NCI)
Collaborator
