Clinical Nomogram Using Novel Computed Tomography-Based Radiomics Predicts Survival in Patients With Non-Small-Cell...

Area of Science:

• Radiomics
• Medical Imaging Analysis
• Computational Topology

Background:

• Accurate survival prediction and risk stratification are crucial for non-small-cell lung cancer (NSCLC) patient management.
• Current methods may not fully capture the prognostic information embedded in tumor characteristics.

Purpose of the Study:

• To introduce and validate the Persistent Homology (PHOM) score, a radiomic quantification of tumor topology, for improved survival prediction in NSCLC.
• To assess the PHOM score's utility in risk stratification for patients undergoing stereotactic body radiation therapy (SBRT).

Main Methods:

• A cohort of 554 patients with stage I or II NSCLC treated with SBRT was analyzed.
• The PHOM score was computed from pretreatment CT scans.
• Multivariable Cox proportional hazards models were used to evaluate OS and cancer-specific survival, incorporating PHOM score and clinical factors.

Main Results:

• The PHOM score was a significant independent predictor of overall survival (HR, 1.17) and cancer-specific survival (HR, 1.31).
• Patients with high PHOM scores exhibited significantly worse median survival (29.2 months) compared to those with low PHOM scores (45.4 months).
• The high-PHOM group had a higher probability of cancer-specific death at 65 months post-treatment.

Conclusions:

• The PHOM score is a valuable radiomic biomarker associated with cancer-specific survival and predictive of overall survival in NSCLC.
• A validated nomogram incorporating the PHOM score can aid in clinical prognosis and post-SBRT treatment decisions.