Journal of Radiation Oncology

This study aimed to (1) compare the agreement of two evaluation methods of metabolic response in patients with non-small cell lung cancer (NSCLC) and determine their prognostic value and (2) explore an optimal cutoff of metabolic reduction to distinguish a more favorable subset of responders.

This is a secondary analysis of prospective studies. Enrolled patients underwent 18F-PET/CT within 2 weeks before, during, and months after radiotherapy (post-RT). Metabolic response was assessed using both Peter MacCallum (PM) method of qualitative visual assessment and University of Michigan (UM) method of semiquantitative measurement. The agreement between two methods determined response, and their prediction of outcome was analyzed.

Forty-four patients with median follow-up of 25.2 months were analyzed. A moderate agreement was observed between PM- and UM-based response assessment (Kappa coefficient = 0.434), unveiling a significant difference in CMR rate (p = 0.001). Categorical responses derived from either method were significantly predictive of overall survival (OS) and progression-free survival (PFS) (p < 0.0001). Numerical percentage decrease of FDG uptake also showed significant correlations with survival, presenting a hazard ratio of 0.97 for both OS and PFS. A 75 % of SUV decrease was found to be the optimal cutoff to predict OS and 2-year progression.

There was a modest discrepancy in metabolic response rates between PM and UM criteria, though both could offer predictive classification for survival. The percentage decrease provides an ordinal value that correlates with prolonged survival, recommending 75 % as the optimal threshold at identifying better responders.