Quantitative Assessment and Predictive Modelling for Treatment Response in Hodgkin's Lymphoma Using 18FDG PET/CT: A Novel Approach

Abstract

Background Assessing treatment response in Hodgkin's lymphoma using 18FDG PET/CT can be challenging due to the nature of false positive of uptaking. This study aims to employ quantitative assessments and create a predictive model for treatment response using newly suggested Criteria in Hodgkin’s lymphoma patients. Methods 62 individuals diagnosed with Hodgkin's lymphoma and treated with chemotherapy were recruited for the research. Up to 6 lesions per patient were selected and delighted for evaluation, and the optimal cutoff was determined using Youden analysis. Predictive parameters for treatment response were identified using the LASSO model, and the new Criteria were evaluated by calibration plot and decision curve analysis (DCA). Results The analysis of 229 lesions led to the development of novel criteria based on the deltaSUVmax, resulting in a NPV of 0.81 and a PPV of 0.86. The LASSO model achieved an AUC of 0.76, with gender, stage, weight, TMV, SUVmaxM, and SUVmeanL identified as significant predictive parameters. Comparative assessment using calibration plots and DCA revealed that the new Criteria delivered more precise outcomes than the conventional visual Criteria. Conclusion Precise evaluation is essential in clinical trials, and continuous efforts are being made to improve the accuracy of response assessment Criteria. Our study found that PET parameters showed a superior specificity to the Deauville Criteria for predicting recurrence/relapse in Hodgkin's lymphoma.