Abstract
BACKGROUND: Integrating immunotherapy with chemotherapy and radiotherapy has revolutionized the treatment of locally advanced non-small cell lung cancer (LA-NSCLC), but challenges persist in accurately assessing treatment responses. This study investigates the use of abbreviated scan protocols for dynamic total body [(18)F]fluorodeoxyglucose positron emission tomography/computed tomography ([(18)F]FDG PET/CT) to enhance clinical feasibility while maintaining diagnostic accuracy for predicting responses to induction immuno-chemotherapy. METHODS: Eligible LA-NSCLC patients scheduled for two cycles of induction immuno-chemotherapy were included and stratified into a testing group and a validation group. In the testing group, three abbreviated dynamic total body [(18)F]FDG PET/CT scan protocols (30-60-minute, 40-60-minute, and 50-60-minute) were developed and compared to the standard scan protocol based on their predictive accuracy of treatment response, as measured by Patlak-Ki values. The optimal abbreviated scan protocol was selected based on its predictive efficiency and scan duration, assessed through receiver operating characteristic (ROC) analysis. The validation group further assessed the predictive performance of the selected scan protocol in relation to treatment response. RESULTS: A total of 38 patients were included in the testing group. The mean bias between the 40-60-minute and standard Patlak-Ki values was closest to zero, indicating minimal error. ROC analysis indicated that the standard Patlak-Ki values had the highest area under the curve (AUC) of 0.75 [95% confidence interval (CI): 0.52-0.97], followed by the 30-60-minute Patlak-Ki values with an AUC of 0.74 (95% CI: 0.52-0.96), the 40-60-minute Patlak-Ki values with an AUC of 0.73 (95% CI: 0.50-0.95), and the 50-60-minute Patlak-Ki values with an AUC of 0.71 (95% CI: 0.50-0.93). Based on the excellent AUC and appropriate scan duration, a 40-60-minute scan protocol was used for a validation group of 34 LA-NSCLC patients before the induction immuno-chemotherapy. The ROC curve for this protocol demonstrated an AUC of 0.78 (95% CI: 0.56-1.00) in predicting treatment response. CONCLUSIONS: The Patlak-Ki values derived from the abbreviated 40-60-minute dynamic scan protocol showed strong agreement with those from the standard dynamic scan protocol in predicting the response to induction immuno-chemotherapy in LA-NSCLC patients. This finding suggested the potential for broader adoption of dynamic total body [(18)F]FDG PET/CT.